{ "video": "https://modelfatigue.news/v/minimax-m3-1/", "numbers": [ { "id": "aar-mm-ii", "what": "Intelligence Index, MiniMax M3.1 Flash", "value": 42.7057475126235, "shown": "42.7", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-cost", "what": "cost per index task at AA's listed prices, MiniMax M3.1 Flash", "value": 0.2201135746720313, "shown": "$0.220", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-cin", "what": "cost per index task spent on input (cached input included), MiniMax M3.1 Flash", "value": 0.17294619922010923, "shown": "$0.173", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-cr", "what": "cost per index task spent on cache reads (part of input), MiniMax M3.1 Flash", "value": 0.12066725957136992, "shown": "$0.121", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-cout", "what": "cost per index task spent on output, MiniMax M3.1 Flash", "value": 0.04716737545192206, "shown": "$0.047", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-out", "what": "output tokens over the whole index run, MiniMax M3.1 Flash", "value": 182566559, "shown": "183M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-reason", "what": "reasoning tokens over the index run (part of output), MiniMax M3.1 Flash", "value": 150758572, "shown": "151M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-in", "what": "input tokens over the index run, MiniMax M3.1 Flash", "value": 8003057663, "shown": "8.0B", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-tpt", "what": "time per index task, MiniMax M3.1 Flash", "value": 768.5950210502266, "shown": "769 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-tb", "what": "Terminal-Bench 4.0, MiniMax M3.1 Flash", "value": 0.257575757575758, "shown": "25.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-tbsci", "what": "Terminal-Bench Science, MiniMax M3.1 Flash", "value": 0.0333333333333333, "shown": "3.3%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-hle", "what": "Humanity's Last Exam, MiniMax M3.1 Flash", "value": 0.479610750695088, "shown": "48.0%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-scicode", "what": "SciCode, MiniMax M3.1 Flash", "value": 0.533564814814815, "shown": "53.4%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-critpt", "what": "CritPt, MiniMax M3.1 Flash", "value": 0.222857142857143, "shown": "22.3%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-lcr", "what": "AA-LCR (long context reasoning), MiniMax M3.1 Flash", "value": 0.796666666666667, "shown": "79.7%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-gdppdf", "what": "GDP.pdf (all pass), MiniMax M3.1 Flash", "value": 0.208, "shown": "20.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-gdpval", "what": "GDPval-AA rating, MiniMax M3.1 Flash", "value": 1457.69, "shown": "1458", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-mmmu", "what": "MMMU Pro, MiniMax M3.1 Flash", "value": 0.791329479768786, "shown": "79.1%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-acc", "what": "AA-Omniscience, share of questions answered right, MiniMax M3.1 Flash", "value": 0.562, "shown": "56.2%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-hall", "what": "AA-Omniscience hallucination rate (wrong answers among the questions it didn't get right), MiniMax M3.1 Flash", "value": 0.928082191780822, "shown": "92.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-mm-omni", "what": "AA-Omniscience Index (right minus wrong), MiniMax M3.1 Flash", "value": 15.55, "shown": "+15.6", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-ii", "what": "Intelligence Index, DeepSeek V4.1 Flash 1017", "value": 45.0965048681912, "shown": "45.1", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-cost", "what": "cost per index task at AA's listed prices, DeepSeek V4.1 Flash 1017", "value": 0.2620484273114172, "shown": "$0.262", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-cin", "what": "cost per index task spent on input (cached input included), DeepSeek V4.1 Flash 1017", "value": 0.15465436463635565, "shown": "$0.155", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-cr", "what": "cost per index task spent on cache reads (part of input), DeepSeek V4.1 Flash 1017", "value": 0.05591718391623929, "shown": "$0.056", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-cout", "what": "cost per index task spent on output, DeepSeek V4.1 Flash 1017", "value": 0.10739406267506157, "shown": "$0.107", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-out", "what": "output tokens over the whole index run, DeepSeek V4.1 Flash 1017", "value": 210859766, "shown": "211M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-reason", "what": "reasoning tokens over the index run (part of output), DeepSeek V4.1 Flash 1017", "value": 170722264, "shown": "171M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-in", "what": "input tokens over the index run, DeepSeek V4.1 Flash 1017", "value": 10279801278, "shown": "10.3B", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-tpt", "what": "time per index task, DeepSeek V4.1 Flash 1017", "value": 357.6048119533332, "shown": "358 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-tb", "what": "Terminal-Bench 4.0, DeepSeek V4.1 Flash 1017", "value": 0.338383838383838, "shown": "33.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-tbsci", "what": "Terminal-Bench Science, DeepSeek V4.1 Flash 1017", "value": 0.128571428571429, "shown": "12.9%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-hle", "what": "Humanity's Last Exam, DeepSeek V4.1 Flash 1017", "value": 0.433271547729379, "shown": "43.3%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-scicode", "what": "SciCode, DeepSeek V4.1 Flash 1017", "value": 0.559027777777778, "shown": "55.9%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-critpt", "what": "CritPt, DeepSeek V4.1 Flash 1017", "value": 0.24, "shown": "24.0%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-lcr", "what": "AA-LCR (long context reasoning), DeepSeek V4.1 Flash 1017", "value": 0.826666666666667, "shown": "82.7%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-gdppdf", "what": "GDP.pdf (all pass), DeepSeek V4.1 Flash 1017", "value": 0.172, "shown": "17.2%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-gdpval", "what": "GDPval-AA rating, DeepSeek V4.1 Flash 1017", "value": 1745.77, "shown": "1746", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-mmmu", "what": "MMMU Pro, DeepSeek V4.1 Flash 1017", "value": 0.80635838150289, "shown": "80.6%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-acc", "what": "AA-Omniscience, share of questions answered right, DeepSeek V4.1 Flash 1017", "value": 0.4285, "shown": "42.9%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-hall", "what": "AA-Omniscience hallucination rate (wrong answers among the questions it didn't get right), DeepSeek V4.1 Flash 1017", "value": 0.6675415573053368, "shown": "66.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-ds-omni", "what": "AA-Omniscience Index (right minus wrong), DeepSeek V4.1 Flash 1017", "value": 4.7, "shown": "+4.7", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-ii", "what": "Intelligence Index, MiniMax M3", "value": 29.2202932236316, "shown": "29.2", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-cost", "what": "cost per index task at AA's listed prices, MiniMax M3", "value": 0.5076126905781733, "shown": "$0.508", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-cin", "what": "cost per index task spent on input (cached input included), MiniMax M3", "value": 0.44978224884613227, "shown": "$0.450", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-cr", "what": "cost per index task spent on cache reads (part of input), MiniMax M3", "value": 0.3556943830256406, "shown": "$0.356", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-cout", "what": "cost per index task spent on output, MiniMax M3", "value": 0.05783044173204086, "shown": "$0.058", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-out", "what": "output tokens over the whole index run, MiniMax M3", "value": 117058967, "shown": "117M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-reason", "what": "reasoning tokens over the index run (part of output), MiniMax M3", "value": 87927151, "shown": "88M", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-in", "what": "input tokens over the index run, MiniMax M3", "value": 5273003361, "shown": "5.3B", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-tpt", "what": "time per index task, MiniMax M3", "value": 519.4905278794248, "shown": "519 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-tb", "what": "Terminal-Bench 4.0, MiniMax M3", "value": 0.0202020202020202, "shown": "2.0%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-tbsci", "what": "Terminal-Bench Science, MiniMax M3", "value": 0.00476190476190476, "shown": "0.5%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-hle", "what": "Humanity's Last Exam, MiniMax M3", "value": 0.389712696941613, "shown": "39.0%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-scicode", "what": "SciCode, MiniMax M3", "value": 0.471064814814815, "shown": "47.1%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-critpt", "what": "CritPt, MiniMax M3", "value": 0.0371428571428571, "shown": "3.7%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-lcr", "what": "AA-LCR (long context reasoning), MiniMax M3", "value": 0.83, "shown": "83.0%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-gdppdf", "what": "GDP.pdf (all pass), MiniMax M3", "value": 0.098, "shown": "9.8%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-gdpval", "what": "GDPval-AA rating, MiniMax M3", "value": 1242.47, "shown": "1242", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-mmmu", "what": "MMMU Pro, MiniMax M3", "value": 0.785549132947977, "shown": "78.6%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-acc", "what": "AA-Omniscience, share of questions answered right, MiniMax M3", "value": 0.167, "shown": "16.7%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-hall", "what": "AA-Omniscience hallucination rate (wrong answers among the questions it didn't get right), MiniMax M3", "value": 0.1842737094837935, "shown": "18.4%", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aar-m3-omni", "what": "AA-Omniscience Index (right minus wrong), MiniMax M3", "value": 1.35, "shown": "+1.4", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-1-out", "what": "output tokens, Pagoda, DeepSeek, run 1", "value": 25527, "shown": "25,527", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-1-reason", "what": "reasoning tokens (part of output), Pagoda, DeepSeek, run 1", "value": 18598, "shown": "18,598", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-1-wall", "what": "time from request to full answer, Pagoda, DeepSeek, run 1", "value": 77.66100239753723, "shown": "78 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-1-route", "what": "route, Pagoda, DeepSeek, run 1", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via InferenceNet", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via InferenceNet", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-2-out", "what": "output tokens, Pagoda, DeepSeek, run 2", "value": 23538, "shown": "23,538", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-2-reason", "what": "reasoning tokens (part of output), Pagoda, DeepSeek, run 2", "value": 16560, "shown": "16,560", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-2-wall", "what": "time from request to full answer, Pagoda, DeepSeek, run 2", "value": 73.05772590637207, "shown": "73 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-deepseek-2-route", "what": "route, Pagoda, DeepSeek, run 2", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via InferenceNet", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via InferenceNet", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-1-out", "what": "output tokens, Pagoda, MiniMax, run 1", "value": 45043, "shown": "45,043", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-1-reason", "what": "reasoning tokens (part of output), Pagoda, MiniMax, run 1", "value": 30949, "shown": "30,949", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-1-wall", "what": "time from request to full answer, Pagoda, MiniMax, run 1", "value": 401.9227421283722, "shown": "402 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-1-route", "what": "route, Pagoda, MiniMax, run 1", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-2-out", "what": "output tokens, Pagoda, MiniMax, run 2", "value": 44199, "shown": "44,199", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-2-reason", "what": "reasoning tokens (part of output), Pagoda, MiniMax, run 2", "value": 29712, "shown": "29,712", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-2-wall", "what": "time from request to full answer, Pagoda, MiniMax, run 2", "value": 375.5822591781616, "shown": "376 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-pagoda-minimax-2-route", "what": "route, Pagoda, MiniMax, run 2", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-1-out", "what": "output tokens, Platformer, DeepSeek, run 1", "value": 35645, "shown": "35,645", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-1-reason", "what": "reasoning tokens (part of output), Platformer, DeepSeek, run 1", "value": 23502, "shown": "23,502", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-1-wall", "what": "time from request to full answer, Platformer, DeepSeek, run 1", "value": 265.91509103775024, "shown": "266 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-1-route", "what": "route, Platformer, DeepSeek, run 1", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-2-out", "what": "output tokens, Platformer, DeepSeek, run 2", "value": 29799, "shown": "29,799", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-2-reason", "what": "reasoning tokens (part of output), Platformer, DeepSeek, run 2", "value": 19999, "shown": "19,999", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-2-wall", "what": "time from request to full answer, Platformer, DeepSeek, run 2", "value": 219.81180596351624, "shown": "220 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-deepseek-2-route", "what": "route, Platformer, DeepSeek, run 2", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-1-out", "what": "output tokens, Platformer, MiniMax, run 1", "value": 41942, "shown": "41,942", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-1-reason", "what": "reasoning tokens (part of output), Platformer, MiniMax, run 1", "value": 30899, "shown": "30,899", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-1-wall", "what": "time from request to full answer, Platformer, MiniMax, run 1", "value": 333.6025528907776, "shown": "334 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-1-route", "what": "route, Platformer, MiniMax, run 1", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-2-out", "what": "output tokens, Platformer, MiniMax, run 2", "value": 48815, "shown": "48,815", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-2-reason", "what": "reasoning tokens (part of output), Platformer, MiniMax, run 2", "value": 38019, "shown": "38,019", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-2-wall", "what": "time from request to full answer, Platformer, MiniMax, run 2", "value": 488.8326120376587, "shown": "489 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-platformer-minimax-2-route", "what": "route, Platformer, MiniMax, run 2", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-1-out", "what": "output tokens, Solar, DeepSeek, run 1", "value": 13869, "shown": "13,869", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-1-reason", "what": "reasoning tokens (part of output), Solar, DeepSeek, run 1", "value": 8839, "shown": "8,839", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-1-wall", "what": "time from request to full answer, Solar, DeepSeek, run 1", "value": 129.11932373046875, "shown": "129 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-1-route", "what": "route, Solar, DeepSeek, run 1", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Wafer", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Wafer", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-2-out", "what": "output tokens, Solar, DeepSeek, run 2", "value": 17109, "shown": "17,109", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-2-reason", "what": "reasoning tokens (part of output), Solar, DeepSeek, run 2", "value": 9546, "shown": "9,546", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-2-wall", "what": "time from request to full answer, Solar, DeepSeek, run 2", "value": 127.57160997390747, "shown": "128 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-deepseek-2-route", "what": "route, Solar, DeepSeek, run 2", "value": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "shown": "OpenRouter, deepseek/deepseek-v4.1-flash-1017 via Relace", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-1-out", "what": "output tokens, Solar, MiniMax, run 1", "value": 41539, "shown": "41,539", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-1-reason", "what": "reasoning tokens (part of output), Solar, MiniMax, run 1", "value": 26370, "shown": "26,370", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-1-wall", "what": "time from request to full answer, Solar, MiniMax, run 1", "value": 348.8581018447876, "shown": "349 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-1-route", "what": "route, Solar, MiniMax, run 1", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-2-out", "what": "output tokens, Solar, MiniMax, run 2", "value": 36275, "shown": "36,275", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-2-reason", "what": "reasoning tokens (part of output), Solar, MiniMax, run 2", "value": 22229, "shown": "22,229", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-2-wall", "what": "time from request to full answer, Solar, MiniMax, run 2", "value": 305.1835820674896, "shown": "305 s", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbsr-solar-minimax-2-route", "what": "route, Solar, MiniMax, run 2", "value": "OpenCode Go, minimax-m3.1-flash-free", "shown": "OpenCode Go, minimax-m3.1-flash-free", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mercury-nasa", "what": "NASA (days over Earth's), Mercury", "value": 0.24096385542168675, "shown": "0.241", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mercury-ds-1", "what": "DeepSeek run 1, as written in its code, Mercury", "value": "0.241", "shown": "0.241", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mercury-ds-2", "what": "DeepSeek run 2, as written in its code, Mercury", "value": "0.2408", "shown": "0.2408", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mercury-mm-1", "what": "MiniMax run 1, as written in its code, Mercury", "value": "0.240846", "shown": "0.240846", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mercury-mm-2", "what": "MiniMax run 2, as written in its code, Mercury", "value": "0.240846", "shown": "0.240846", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-venus-nasa", "what": "NASA (days over Earth's), Venus", "value": 0.6152792990142387, "shown": "0.615", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-venus-ds-1", "what": "DeepSeek run 1, as written in its code, Venus", "value": "0.615", "shown": "0.615", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-venus-ds-2", "what": "DeepSeek run 2, as written in its code, Venus", "value": "0.6152", "shown": "0.6152", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-venus-mm-1", "what": "MiniMax run 1, as written in its code, Venus", "value": "0.615197", "shown": "0.615197", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-venus-mm-2", "what": "MiniMax run 2, as written in its code, Venus", "value": "0.615197", "shown": "0.615197", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-earth-nasa", "what": "NASA (days over Earth's), Earth", "value": 1.0, "shown": "1.000", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-earth-ds-1", "what": "DeepSeek run 1, as written in its code, Earth", "value": "1", "shown": "1", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-earth-ds-2", "what": "DeepSeek run 2, as written in its code, Earth", "value": "1", "shown": "1", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-earth-mm-1", "what": "MiniMax run 1, as written in its code, Earth", "value": "1", "shown": "1", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-earth-mm-2", "what": "MiniMax run 2, as written in its code, Earth", "value": "1", "shown": "1", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mars-nasa", "what": "NASA (days over Earth's), Mars", "value": 1.8811610076670318, "shown": "1.881", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mars-ds-1", "what": "DeepSeek run 1, as written in its code, Mars", "value": "1.881", "shown": "1.881", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mars-ds-2", "what": "DeepSeek run 2, as written in its code, Mars", "value": "1.8808", "shown": "1.8808", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mars-mm-1", "what": "MiniMax run 1, as written in its code, Mars", "value": "1.8808", "shown": "1.8808", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-mars-mm-2", "what": "MiniMax run 2, as written in its code, Mars", "value": "1.8808", "shown": "1.8808", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-jupiter-nasa", "what": "NASA (days over Earth's), Jupiter", "value": 11.859255202628697, "shown": "11.859", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-jupiter-ds-1", "what": "DeepSeek run 1, as written in its code, Jupiter", "value": "11.86", "shown": "11.86", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-jupiter-ds-2", "what": "DeepSeek run 2, as written in its code, Jupiter", "value": "11.862", "shown": "11.862", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-jupiter-mm-1", "what": "MiniMax run 1, as written in its code, Jupiter", "value": "11.862", "shown": "11.862", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-jupiter-mm-2", "what": "MiniMax run 2, as written in its code, Jupiter", "value": "11.862", "shown": "11.862", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-saturn-nasa", "what": "NASA (days over Earth's), Saturn", "value": 29.427710843373497, "shown": "29.428", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-saturn-ds-1", "what": "DeepSeek run 1, as written in its code, Saturn", "value": "29.45", "shown": "29.45", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-saturn-ds-2", "what": "DeepSeek run 2, as written in its code, Saturn", "value": "29.457", "shown": "29.457", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-saturn-mm-1", "what": "MiniMax run 1, as written in its code, Saturn", "value": "29.457", "shown": "29.457", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-saturn-mm-2", "what": "MiniMax run 2, as written in its code, Saturn", "value": "29.457", "shown": "29.457", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-uranus-nasa", "what": "NASA (days over Earth's), Uranus", "value": 83.75958378970428, "shown": "83.760", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-uranus-ds-1", "what": "DeepSeek run 1, as written in its code, Uranus", "value": "84.02", "shown": "84.02", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-uranus-ds-2", "what": "DeepSeek run 2, as written in its code, Uranus", "value": "84.02", "shown": "84.02", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-uranus-mm-1", "what": "MiniMax run 1, as written in its code, Uranus", "value": "84.011", "shown": "84.011", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-uranus-mm-2", "what": "MiniMax run 2, as written in its code, Uranus", "value": "84.011", "shown": "84.011", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-neptune-nasa", "what": "NASA (days over Earth's), Neptune", "value": 163.74589266155533, "shown": "163.746", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-neptune-ds-1", "what": "DeepSeek run 1, as written in its code, Neptune", "value": "164.8", "shown": "164.8", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-neptune-ds-2", "what": "DeepSeek run 2, as written in its code, Neptune", "value": "164.79", "shown": "164.79", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-neptune-mm-1", "what": "MiniMax run 1, as written in its code, Neptune", "value": "164.79", "shown": "164.79", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-neptune-mm-2", "what": "MiniMax run 2, as written in its code, Neptune", "value": "164.79", "shown": "164.79", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-1-wall", "what": "time from request to full answer, Gears, MiniMax, run 1", "value": 321.7, "shown": "322 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-1-out", "what": "output tokens, Gears, MiniMax, run 1", "value": 26048, "shown": "26,048", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-1-speeds", "what": "gears whose speed was right, of five, Gears, MiniMax, run 1", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-1-dirs", "what": "gears turning the right way, of five, Gears, MiniMax, run 1", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-1-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, MiniMax, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-2-wall", "what": "time from request to full answer, Gears, MiniMax, run 2", "value": 309.1, "shown": "309 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-2-out", "what": "output tokens, Gears, MiniMax, run 2", "value": 33925, "shown": "33,925", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-2-speeds", "what": "gears whose speed was right, of five, Gears, MiniMax, run 2", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-2-dirs", "what": "gears turning the right way, of five, Gears, MiniMax, run 2", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-2-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, MiniMax, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-3-wall", "what": "time from request to full answer, Gears, MiniMax, run 3", "value": 272.2, "shown": "272 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-3-out", "what": "output tokens, Gears, MiniMax, run 3", "value": 24626, "shown": "24,626", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-3-speeds", "what": "gears whose speed was right, of five, Gears, MiniMax, run 3", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-3-dirs", "what": "gears turning the right way, of five, Gears, MiniMax, run 3", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-mm-3-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, MiniMax, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-1-wall", "what": "time from request to full answer, Gears, DeepSeek, run 1", "value": 111.7, "shown": "112 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-1-out", "what": "output tokens, Gears, DeepSeek, run 1", "value": 28976, "shown": "28,976", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-1-speeds", "what": "gears whose speed was right, of five, Gears, DeepSeek, run 1", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-1-dirs", "what": "gears turning the right way, of five, Gears, DeepSeek, run 1", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-1-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, DeepSeek, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-2-wall", "what": "time from request to full answer, Gears, DeepSeek, run 2", "value": 147.0, "shown": "147 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-2-out", "what": "output tokens, Gears, DeepSeek, run 2", "value": 34113, "shown": "34,113", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-2-speeds", "what": "gears whose speed was right, of five, Gears, DeepSeek, run 2", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-2-dirs", "what": "gears turning the right way, of five, Gears, DeepSeek, run 2", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-2-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, DeepSeek, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-3-wall", "what": "time from request to full answer, Gears, DeepSeek, run 3", "value": 94.9, "shown": "95 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-3-out", "what": "output tokens, Gears, DeepSeek, run 3", "value": 20762, "shown": "20,762", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-3-speeds", "what": "gears whose speed was right, of five, Gears, DeepSeek, run 3", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-3-dirs", "what": "gears turning the right way, of five, Gears, DeepSeek, run 3", "value": 5, "shown": "5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-gears-ds-3-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Gears, DeepSeek, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-1-wall", "what": "time from request to full answer, Planetary, MiniMax, run 1", "value": 381.0, "shown": "381 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-1-out", "what": "output tokens, Planetary, MiniMax, run 1", "value": 41008, "shown": "41,008", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-1-carrier", "what": "carrier speed read off the page, rpm, Planetary, MiniMax, run 1", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-1-planet", "what": "planet gears' speed read off the page, rpm, Planetary, MiniMax, run 1", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-1-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, MiniMax, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-2-wall", "what": "time from request to full answer, Planetary, MiniMax, run 2", "value": 423.0, "shown": "423 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-2-out", "what": "output tokens, Planetary, MiniMax, run 2", "value": 47506, "shown": "47,506", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-2-carrier", "what": "carrier speed read off the page, rpm, Planetary, MiniMax, run 2", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-2-planet", "what": "planet gears' speed read off the page, rpm, Planetary, MiniMax, run 2", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-2-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, MiniMax, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-3-wall", "what": "time from request to full answer, Planetary, MiniMax, run 3", "value": 280.4, "shown": "280 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-3-out", "what": "output tokens, Planetary, MiniMax, run 3", "value": 32159, "shown": "32,159", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-3-carrier", "what": "carrier speed read off the page, rpm, Planetary, MiniMax, run 3", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-3-planet", "what": "planet gears' speed read off the page, rpm, Planetary, MiniMax, run 3", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-mm-3-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, MiniMax, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-1-wall", "what": "time from request to full answer, Planetary, DeepSeek, run 1", "value": 193.0, "shown": "193 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-1-out", "what": "output tokens, Planetary, DeepSeek, run 1", "value": 47210, "shown": "47,210", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-1-carrier", "what": "carrier speed read off the page, rpm, Planetary, DeepSeek, run 1", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-1-planet", "what": "planet gears' speed read off the page, rpm, Planetary, DeepSeek, run 1", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-1-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, DeepSeek, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-2-wall", "what": "time from request to full answer, Planetary, DeepSeek, run 2", "value": 192.3, "shown": "192 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-2-out", "what": "output tokens, Planetary, DeepSeek, run 2", "value": 43796, "shown": "43,796", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-2-carrier", "what": "carrier speed read off the page, rpm, Planetary, DeepSeek, run 2", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-2-planet", "what": "planet gears' speed read off the page, rpm, Planetary, DeepSeek, run 2", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-2-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, DeepSeek, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-3-wall", "what": "time from request to full answer, Planetary, DeepSeek, run 3", "value": 222.4, "shown": "222 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-3-out", "what": "output tokens, Planetary, DeepSeek, run 3", "value": 48228, "shown": "48,228", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-3-carrier", "what": "carrier speed read off the page, rpm, Planetary, DeepSeek, run 3", "value": 17.14, "shown": "17.14", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-3-planet", "what": "planet gears' speed read off the page, rpm, Planetary, DeepSeek, run 3", "value": -40.0, "shown": "−40", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-planetary-ds-3-mesh", "what": "teeth meshing, scored blind from a still frame (1 to 3), Planetary, DeepSeek, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-wall", "what": "time from request to full answer, Cradle, MiniMax, run 1", "value": 412.9, "shown": "413 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-out", "what": "output tokens, Cradle, MiniMax, run 1", "value": 29634, "shown": "29,634", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-presses", "what": "presses that sent out the right number of balls, of three, Cradle, MiniMax, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-o1", "what": "balls out on the right after pressing 1, Cradle, MiniMax, run 1", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-o2", "what": "balls out on the right after pressing 2, Cradle, MiniMax, run 1", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-1-o3", "what": "balls out on the right after pressing 3, Cradle, MiniMax, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-wall", "what": "time from request to full answer, Cradle, MiniMax, run 2", "value": 290.3, "shown": "290 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-out", "what": "output tokens, Cradle, MiniMax, run 2", "value": 32964, "shown": "32,964", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-presses", "what": "presses that sent out the right number of balls, of three, Cradle, MiniMax, run 2", "value": 0, "shown": "0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-o1", "what": "balls out on the right after pressing 1, Cradle, MiniMax, run 2", "value": 0, "shown": "0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-o2", "what": "balls out on the right after pressing 2, Cradle, MiniMax, run 2", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-2-o3", "what": "balls out on the right after pressing 3, Cradle, MiniMax, run 2", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-wall", "what": "time from request to full answer, Cradle, MiniMax, run 3", "value": 918.6, "shown": "919 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-out", "what": "output tokens, Cradle, MiniMax, run 3", "value": 26304, "shown": "26,304", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-presses", "what": "presses that sent out the right number of balls, of three, Cradle, MiniMax, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-o1", "what": "balls out on the right after pressing 1, Cradle, MiniMax, run 3", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-o2", "what": "balls out on the right after pressing 2, Cradle, MiniMax, run 3", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-mm-3-o3", "what": "balls out on the right after pressing 3, Cradle, MiniMax, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-wall", "what": "time from request to full answer, Cradle, DeepSeek, run 1", "value": 86.7, "shown": "87 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-out", "what": "output tokens, Cradle, DeepSeek, run 1", "value": 20629, "shown": "20,629", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-presses", "what": "presses that sent out the right number of balls, of three, Cradle, DeepSeek, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-o1", "what": "balls out on the right after pressing 1, Cradle, DeepSeek, run 1", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-o2", "what": "balls out on the right after pressing 2, Cradle, DeepSeek, run 1", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-1-o3", "what": "balls out on the right after pressing 3, Cradle, DeepSeek, run 1", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-wall", "what": "time from request to full answer, Cradle, DeepSeek, run 2", "value": 86.9, "shown": "87 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-out", "what": "output tokens, Cradle, DeepSeek, run 2", "value": 20550, "shown": "20,550", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-presses", "what": "presses that sent out the right number of balls, of three, Cradle, DeepSeek, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-o1", "what": "balls out on the right after pressing 1, Cradle, DeepSeek, run 2", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-o2", "what": "balls out on the right after pressing 2, Cradle, DeepSeek, run 2", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-2-o3", "what": "balls out on the right after pressing 3, Cradle, DeepSeek, run 2", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-wall", "what": "time from request to full answer, Cradle, DeepSeek, run 3", "value": 117.4, "shown": "117 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-out", "what": "output tokens, Cradle, DeepSeek, run 3", "value": 29004, "shown": "29,004", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-presses", "what": "presses that sent out the right number of balls, of three, Cradle, DeepSeek, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-o1", "what": "balls out on the right after pressing 1, Cradle, DeepSeek, run 3", "value": 1, "shown": "1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-o2", "what": "balls out on the right after pressing 2, Cradle, DeepSeek, run 3", "value": 2, "shown": "2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-cradle-ds-3-o3", "what": "balls out on the right after pressing 3, Cradle, DeepSeek, run 3", "value": 3, "shown": "3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-1-wall", "what": "time from request to full answer, Moon, MiniMax, run 1", "value": 1139.0, "shown": "1139 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-1-out", "what": "output tokens, Moon, MiniMax, run 1", "value": 56360, "shown": "56,360", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-1-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, MiniMax, run 1", "value": 0.1, "shown": "0.1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-1-sun", "what": "drawn Sun's error, degrees, Moon, MiniMax, run 1", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-1-illum", "what": "illuminated fraction it gave for the Moon, Moon, MiniMax, run 1", "value": 0.6264, "shown": "0.626", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-2-wall", "what": "time from request to full answer, Moon, MiniMax, run 2", "value": 439.8, "shown": "440 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-2-out", "what": "output tokens, Moon, MiniMax, run 2", "value": 46319, "shown": "46,319", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-2-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, MiniMax, run 2", "value": 97.5, "shown": "97.5", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-2-sun", "what": "drawn Sun's error, degrees, Moon, MiniMax, run 2", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-2-illum", "what": "illuminated fraction it gave for the Moon, Moon, MiniMax, run 2", "value": 0.6264, "shown": "0.626", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-3-wall", "what": "time from request to full answer, Moon, MiniMax, run 3", "value": 1042.6, "shown": "1043 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-3-out", "what": "output tokens, Moon, MiniMax, run 3", "value": 58115, "shown": "58,115", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-3-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, MiniMax, run 3", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-3-sun", "what": "drawn Sun's error, degrees, Moon, MiniMax, run 3", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-mm-3-illum", "what": "illuminated fraction it gave for the Moon, Moon, MiniMax, run 3", "value": 0.6259, "shown": "0.626", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-1-wall", "what": "time from request to full answer, Moon, DeepSeek, run 1", "value": 142.2, "shown": "142 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-1-out", "what": "output tokens, Moon, DeepSeek, run 1", "value": 34554, "shown": "34,554", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-1-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, DeepSeek, run 1", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-1-sun", "what": "drawn Sun's error, degrees, Moon, DeepSeek, run 1", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-1-illum", "what": "illuminated fraction it gave for the Moon, Moon, DeepSeek, run 1", "value": 0.62719, "shown": "0.627", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-2-wall", "what": "time from request to full answer, Moon, DeepSeek, run 2", "value": 183.7, "shown": "184 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-2-out", "what": "output tokens, Moon, DeepSeek, run 2", "value": 45908, "shown": "45,908", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-2-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, DeepSeek, run 2", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-2-sun", "what": "drawn Sun's error, degrees, Moon, DeepSeek, run 2", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-2-illum", "what": "illuminated fraction it gave for the Moon, Moon, DeepSeek, run 2", "value": 0.627, "shown": "0.627", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-3-wall", "what": "time from request to full answer, Moon, DeepSeek, run 3", "value": 158.6, "shown": "159 s", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-3-out", "what": "output tokens, Moon, DeepSeek, run 3", "value": 42311, "shown": "42,311", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-3-moon", "what": "drawn Moon's error against JPL DE440s, degrees, Moon, DeepSeek, run 3", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-3-sun", "what": "drawn Sun's error, degrees, Moon, DeepSeek, run 3", "value": 0.0, "shown": "0.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vtr-moon-ds-3-illum", "what": "illuminated fraction it gave for the Moon, Moon, DeepSeek, run 3", "value": 0.627014, "shown": "0.627", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mm-in", "what": "MiniMax M3.1 Flash, price per million input tokens, at all hours", "value": 0.15, "shown": "$0.15", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mm-out", "what": "MiniMax M3.1 Flash, price per million output tokens, at all hours", "value": 0.6, "shown": "$0.60", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mm-cache", "what": "MiniMax M3.1 Flash, price per million cached input tokens", "value": 0.015, "shown": "$0.015", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-in", "what": "DeepSeek V4.1 Flash 1017, price per million input tokens (cache miss), off-peak", "value": 0.15, "shown": "$0.15", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-out", "what": "DeepSeek V4.1 Flash 1017, price per million output tokens, off-peak", "value": 0.6, "shown": "$0.60", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cache", "what": "DeepSeek V4.1 Flash 1017, price per million cached input tokens (cache hit), off-peak", "value": 0.003, "shown": "$0.003", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-in-peak", "what": "DeepSeek V4.1 Flash 1017, price per million input tokens (cache miss), peak", "value": 0.3, "shown": "$0.30", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-out-peak", "what": "DeepSeek V4.1 Flash 1017, price per million output tokens, peak", "value": 1.2, "shown": "$1.20", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cache-peak", "what": "DeepSeek V4.1 Flash 1017, price per million cached input tokens, peak", "value": 0.006, "shown": "$0.006", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-hours", "what": "DeepSeek's peak hours, Monday to Friday, except Chinese public holidays", "value": "01:00–04:00 and 06:00–10:00 UTC", "shown": "01:00–04:00 and 06:00–10:00 UTC", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-half", "what": "DeepSeek's off-peak rates as a share of its peak rates", "value": 0.5, "shown": "50%", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "m3-in", "what": "MiniMax M3, price per million input tokens (prompts up to 512K tokens)", "value": 0.3, "shown": "$0.30", "whose": "MiniMax", "source": "MiniMax, API pricing overview", "url": "https://platform.minimax.io/docs/pricing/overview", "read": "2026-10-19T17:10", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "m3-out", "what": "MiniMax M3, price per million output tokens (prompts up to 512K tokens)", "value": 1.2, "shown": "$1.20", "whose": "MiniMax", "source": "MiniMax, API pricing overview", "url": "https://platform.minimax.io/docs/pricing/overview", "read": "2026-10-19T17:10", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "m3-tier", "what": "the prompt length M3's quoted price applies to", "value": "up to 512K input tokens", "shown": "up to 512K input tokens", "whose": "MiniMax", "source": "MiniMax, API pricing overview", "url": "https://platform.minimax.io/docs/pricing/overview", "read": "2026-10-19T17:10", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "effort-top", "what": "M3.1 Flash's two highest reasoning-effort settings (the last two of five)", "value": "xhigh and max", "shown": "xhigh and max", "whose": "MiniMax", "source": "MiniMax, MiniMax-M3.1-Flash model card on Hugging Face (its benchmark table and model details)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-ds-pin", "what": "the input price Artificial Analysis lists for DeepSeek V4.1 Flash 1017, per million tokens", "value": 0.3, "shown": "$0.30", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-ds-pout", "what": "the output price Artificial Analysis lists for DeepSeek V4.1 Flash 1017, per million tokens", "value": 1.2, "shown": "$1.20", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-mm-pin", "what": "the input price Artificial Analysis lists for MiniMax M3.1 Flash, per million tokens", "value": 0.15, "shown": "$0.15", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-mm-pout", "what": "the output price Artificial Analysis lists for MiniMax M3.1 Flash, per million tokens", "value": 0.6, "shown": "$0.60", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-m3-pin", "what": "the input price Artificial Analysis lists for MiniMax M3, per million tokens", "value": 0.3, "shown": "$0.30", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-m3-pout", "what": "the output price Artificial Analysis lists for MiniMax M3, per million tokens", "value": 1.2, "shown": "$1.20", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-ctx", "what": "DeepSeek V4.1 Flash 1017's context length on DeepSeek's API (deepseek-flash)", "value": 1000000.0, "shown": "1M", "whose": "DeepSeek", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "or-paid", "what": "paid endpoints OpenRouter lists for MiniMax M3.1 Flash", "value": 12, "shown": "12", "whose": "OpenRouter", "source": "OpenRouter, MiniMax M3.1 Flash endpoints", "url": "https://openrouter.ai/minimax/minimax-m3.1-flash/providers", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "or-free", "what": "endpoints OpenRouter lists for the free variant of MiniMax M3.1 Flash", "value": 0, "shown": "0", "whose": "OpenRouter", "source": "OpenRouter, MiniMax M3.1 Flash (free) endpoints", "url": "https://openrouter.ai/minimax/minimax-m3.1-flash:free/providers", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "bunny", "what": "what Space Bunny processed while it ran anonymously (MiniMax's words)", "value": "more than 200 trillion tokens", "shown": "more than 200 trillion tokens", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "claim", "what": "MiniMax's own claim for M3.1 Flash (not a measurement)", "value": "nearly three times faster and reducing usage costs by 70%", "shown": "nearly three times faster and reducing usage costs by 70%", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "claim-70", "what": "MiniMax: how much lower M3.1 Flash's overall usage cost is than M3's, for agentic tasks", "value": 0.7, "shown": "70%", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "claim-tps", "what": "MiniMax's own claim for M3.1 Flash's output speed in everyday use", "value": "200+ TPS, with a median (P50) of 170+ TPS", "shown": "200+ TPS, with a median (P50) of 170+ TPS", "whose": "MiniMax", "source": "MiniMax, “MiniMax M3.1 Flash: Redefining Agent Performance and Cost” (release post, 19 October 2026)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ctx", "what": "MiniMax M3.1 Flash's context window, tokens", "value": 1000000.0, "shown": "1M", "whose": "MiniMax", "source": "MiniMax, MiniMax-M3.1-Flash model card on Hugging Face (its benchmark table and model details)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "params", "what": "MiniMax M3.1 Flash's size, as MiniMax gives it", "value": "about 428B in total, 23B active per token", "shown": "about 428B in total, 23B active per token", "whose": "MiniMax", "source": "MiniMax, MiniMax-M3.1-Flash model card on Hugging Face (its benchmark table and model details)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "hf-modified", "what": "the time of the weights repository's last change, on 23 October", "value": "03:37 UTC", "shown": "03:37 UTC", "whose": "Hugging Face", "source": "Hugging Face, the MiniMaxAI/MiniMax-M3.1-Flash repository (public, not gated)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "weights-date", "what": "the date MiniMax gave on 19 October for the weights", "value": "Oct 23", "shown": "Oct 23", "whose": "MiniMax", "source": "MiniMax (@MiniMax_AI) on X, 19 October 2026, the launch post", "url": "https://x.com/MiniMax_AI/status/2112150867868541221", "read": "2026-10-19T15:50", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lic-attr", "what": "the line commercial users must display", "value": "Built with MiniMax M3.1-Flash", "shown": "Built with MiniMax M3.1-Flash", "whose": "MiniMax", "source": "MiniMax Community License (the licence file in the weights' repository)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash/blob/main/LICENSE", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lic-rev", "what": "yearly revenue above which commercial use needs MiniMax's prior written authorisation", "value": "20 million US dollars", "shown": "20 million US dollars", "whose": "MiniMax", "source": "MiniMax Community License (the licence file in the weights' repository)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash/blob/main/LICENSE", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-ds-name", "what": "the name of DeepSeek's row on Artificial Analysis's page", "value": "DeepSeek V4.1 Flash 1017 (Max)", "shown": "DeepSeek V4.1 Flash 1017 (Max)", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-tests", "what": "tests in the Intelligence Index", "value": 10, "shown": "10", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (the index's description)", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-ds-gdppdf-19", "what": "GDP.pdf (all pass) on Artificial Analysis, DeepSeek V4.1 Flash 1017, as read on 19 October", "value": 0.172, "shown": "17.2%", "whose": "Artificial Analysis", "source": "Artificial Analysis, DeepSeek V4.1 Flash 1017 model page, re-read on 19 October", "url": "https://artificialanalysis.ai/models/deepseek-v4-1-flash-1017", "read": "2026-10-19T07:46", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa5-ver", "what": "the version of Artificial Analysis's Intelligence Index on 7 November (v5.0.1)", "value": 5, "shown": "5", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash and DeepSeek V4.1 Flash 1017 model pages on Intelligence Index v5.0.1, read 7 November", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-11-07T16:43", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa5-mm-omni", "what": "AA-Omniscience Index on Intelligence Index v5.0.1, MiniMax M3.1 Flash, read 7 November", "value": -6.4, "shown": "-6.4", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash and DeepSeek V4.1 Flash 1017 model pages on Intelligence Index v5.0.1, read 7 November", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-11-07T16:43", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa5-ds-omni", "what": "AA-Omniscience Index on Intelligence Index v5.0.1, DeepSeek V4.1 Flash 1017, read 7 November", "value": 0.9, "shown": "0.9", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash and DeepSeek V4.1 Flash 1017 model pages on Intelligence Index v5.0.1, read 7 November", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-11-07T16:43", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-mm-ii-19", "what": "MiniMax M3.1 Flash's Intelligence Index on release day", "value": 42.3811052132362, "shown": "42.4", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page, release day", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "r-mm", "what": "AA-Omniscience, MiniMax M3.1 Flash: share of all questions not answered, or only partly answered: what is left of 100% once the right and wrong shares, each rounded to a tenth of a point, are taken away (our arithmetic, as the video draws it)", "value": 0.03199999999999996, "shown": "3.2%", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "r-ds", "what": "AA-Omniscience, DeepSeek V4.1 Flash 1017: share of all questions not answered, or only partly answered: what is left of 100% once the right and wrong shares, each rounded to a tenth of a point, are taken away (our arithmetic, as the video draws it)", "value": 0.19, "shown": "19.0%", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-mm-19", "what": "median output speed, MiniMax M3.1 Flash, tokens per second, 19 October", "value": 122.805319930525, "shown": "122.8", "whose": "Artificial Analysis", "source": "Artificial Analysis, median output speed on 19 October (each model's own object in AA's saved Mercury 2.5 page)", "url": "https://artificialanalysis.ai/models/mercury-2-5", "read": "2026-10-19T20:21", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-mm-21", "what": "median output speed, MiniMax M3.1 Flash, tokens per second, 21 October", "value": 93.1119234389769, "shown": "93.1", "whose": "Artificial Analysis", "source": "Artificial Analysis, median output speed on 21 October (our re-read of the M3.1 Flash page)", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-21T09:18", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-mm", "what": "median output speed, MiniMax M3.1 Flash, tokens per second, 25 October", "value": 95.5531649197315, "shown": "95.6", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "fat-mm", "what": "time to the first answer token after reasoning, long prompt, MiniMax M3.1 Flash, 25 October", "value": 22.246030717698474, "shown": "22.2 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ttfc-mm", "what": "time to the first token (reasoning included), MiniMax M3.1 Flash, 25 October", "value": 1.31527450800013, "shown": "1.32 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-ds-19", "what": "median output speed, DeepSeek V4.1 Flash 1017, tokens per second, 19 October", "value": 216.482278849785, "shown": "216.5", "whose": "Artificial Analysis", "source": "Artificial Analysis, median output speed on 19 October (each model's own object in AA's saved Mercury 2.5 page)", "url": "https://artificialanalysis.ai/models/mercury-2-5", "read": "2026-10-19T20:21", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-ds-21", "what": "median output speed, DeepSeek V4.1 Flash 1017, tokens per second, 21 October", "value": 208.679204176946, "shown": "208.7", "whose": "Artificial Analysis", "source": "Artificial Analysis, median output speed on 21 October (our re-read of the M3.1 Flash page)", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-21T09:18", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-ds", "what": "median output speed, DeepSeek V4.1 Flash 1017, tokens per second, 25 October", "value": 195.157086034074, "shown": "195.2", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "fat-ds", "what": "time to the first answer token after reasoning, long prompt, DeepSeek V4.1 Flash 1017, 25 October", "value": 11.390313256556013, "shown": "11.4 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ttfc-ds", "what": "time to the first token (reasoning included), DeepSeek V4.1 Flash 1017, 25 October", "value": 1.14215859999999, "shown": "1.14 s", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "spd-m3", "what": "median output speed, MiniMax M3, tokens per second, 25 October", "value": 88.7245235883505, "shown": "88.7", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-auto-mm", "what": "MiniMax's table, Automation-Bench v1.0.6, MiniMax M3.1 Flash", "value": 62.8, "shown": "62.8", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-auto-ds", "what": "MiniMax's table, Automation-Bench v1.0.6, DeepSeek V4.1 Flash 1017", "value": 60.5, "shown": "60.5", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-banker-mm", "what": "MiniMax's table, BankerToolBench, MiniMax M3.1 Flash", "value": 76.2, "shown": "76.2", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-gdppdf-mm", "what": "MiniMax's table, GDP.pdf AA, MiniMax M3.1 Flash", "value": 20.8, "shown": "20.8", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-tb-mm", "what": "MiniMax's table, Terminal Bench 4.0, MiniMax M3.1 Flash", "value": 38.1, "shown": "38.1", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-tb-ds", "what": "MiniMax's table, Terminal Bench 4.0, DeepSeek V4.1 Flash 1017", "value": 42.3, "shown": "42.3", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-loca-mm", "what": "MiniMax's table, LOCA-Bench 256K, MiniMax M3.1 Flash", "value": 60.3, "shown": "60.3", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-ale-mm", "what": "MiniMax's table, Agents’ Last Exam, MiniMax M3.1 Flash", "value": 29.6, "shown": "29.6", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-osworld-mm", "what": "MiniMax's table, OSWorld 2.0, MiniMax M3.1 Flash", "value": 61.7, "shown": "61.7", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-posttrain-mm", "what": "MiniMax's table, PostTrain Bench v1.1, MiniMax M3.1 Flash", "value": 39.3, "shown": "39.3", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-deepswe-mm", "what": "MiniMax's table, DeepSWE v1.1, MiniMax M3.1 Flash", "value": 73.7, "shown": "73.7", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-deepswe-ds", "what": "MiniMax's table, DeepSWE v1.1, DeepSeek V4.1 Flash 1017", "value": 77.0, "shown": "77.0", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-nl2repo-mm", "what": "MiniMax's table, NL2Repo-Bench, MiniMax M3.1 Flash", "value": 61.2, "shown": "61.2", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-nl2repo-ds", "what": "MiniMax's table, NL2Repo-Bench, DeepSeek V4.1 Flash 1017", "value": 68.2, "shown": "68.2", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-refactor-mm", "what": "MiniMax's table, SWE Refactor Bench, MiniMax M3.1 Flash", "value": 23.0, "shown": "23.0", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-cybergym-mm", "what": "MiniMax's table, CyberGym, MiniMax M3.1 Flash", "value": 82.2, "shown": "82.2", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-babyvision-mm", "what": "MiniMax's table, BabyVision (w/tools), MiniMax M3.1 Flash", "value": 91.0, "shown": "91.0", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-babyvision-ds", "what": "MiniMax's table, BabyVision (w/tools), DeepSeek V4.1 Flash 1017", "value": 92.5, "shown": "92.5", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-hle-mm", "what": "MiniMax's table, HLE (w/o tools), MiniMax M3.1 Flash", "value": 43.4, "shown": "43.4", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-hle-ds", "what": "MiniMax's table, HLE (w/o tools), DeepSeek V4.1 Flash 1017", "value": 42.6, "shown": "42.6", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-charxiv-mm", "what": "MiniMax's table, CharXiv RQ, MiniMax M3.1 Flash", "value": 82.5, "shown": "82.5", "whose": "MiniMax", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmc-hle-ds", "what": "the Hugging Face card's table, HLE (without tools), DeepSeek V4.1 Flash 1017", "value": 42.6, "shown": "42.6", "whose": "MiniMax", "source": "MiniMax, the benchmark table on the Hugging Face model card (parsed from its HTML)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmb-hle-ds", "what": "the release post's table of 19 October, HLE (without tools), DeepSeek V4.1 Flash 1017", "value": 44.0, "shown": "44.0", "whose": "MiniMax", "source": "MiniMax, the same table in the release post of 19 October (an image)", "url": "https://www.minimax.io/blog/minimax-m31f", "read": "2026-10-19T17:09", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "aa-hle-n", "what": "questions in Artificial Analysis's run of Humanity's Last Exam", "value": 2158, "shown": "2,158", "whose": "Artificial Analysis", "source": "Artificial Analysis, Intelligence Benchmarking methodology (AA-Omniscience's scoring, Humanity's Last Exam)", "url": "https://artificialanalysis.ai/methodology/intelligence-benchmarking", "read": "2026-10-01", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmc-hle-n", "what": "questions in the text-only part of HLE MiniMax ran", "value": 2158, "shown": "2,158", "whose": "MiniMax", "source": "MiniMax, MiniMax-M3.1-Flash model card on Hugging Face (its benchmark table and model details)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmc-tb-h", "what": "the harness MiniMax ran Terminal-Bench 4.0 in", "value": "Codex 0.149.1", "shown": "Codex 0.149.1", "whose": "MiniMax", "source": "MiniMax, MiniMax-M3.1-Flash model card on Hugging Face (its benchmark table and model details)", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "dsc-tb", "what": "DeepSeek's card, Terminal-Bench 4.0, DeepSeek V4.1 Flash 1017 (Claude Code harness)", "value": 42.3, "shown": "42.3", "whose": "DeepSeek", "source": "DeepSeek, DeepSeek-V4.1-Flash-1017 model card on Hugging Face", "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017", "read": "2026-10-17T23:00", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "dsc-ale", "what": "DeepSeek's card, Agents' Last Exam, DeepSeek V4.1 Flash 1017", "value": 35.4, "shown": "35.4", "whose": "DeepSeek", "source": "DeepSeek, DeepSeek-V4.1-Flash-1017 model card on Hugging Face", "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017", "read": "2026-10-17T23:00", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "dsc-hle", "what": "DeepSeek's card, HLE without tools, DeepSeek V4.1 Flash 1017, the full set", "value": 42.6, "shown": "42.6", "whose": "DeepSeek", "source": "DeepSeek, DeepSeek-V4.1-Flash-1017 model card on Hugging Face", "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017", "read": "2026-10-17T23:00", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "dsc-hle-t", "what": "DeepSeek's card, HLE without tools, DeepSeek V4.1 Flash 1017, the text-only subset", "value": 44.0, "shown": "44.0", "whose": "DeepSeek", "source": "DeepSeek, DeepSeek-V4.1-Flash-1017 model card on Hugging Face", "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017", "read": "2026-10-17T23:00", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "orq-inet", "what": "the quantization OpenRouter lists for InferenceNet's DeepSeek V4.1 Flash 1017 endpoint", "value": "fp8", "shown": "fp8", "whose": "OpenRouter", "source": "OpenRouter, DeepSeek V4.1 Flash 1017 endpoints", "url": "https://openrouter.ai/deepseek/deepseek-v4.1-flash-20261017/providers", "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "orq-relace", "what": "the quantization OpenRouter lists for Relace's DeepSeek V4.1 Flash 1017 endpoint", "value": "unknown", "shown": "unknown", "whose": "OpenRouter", "source": "OpenRouter, DeepSeek V4.1 Flash 1017 endpoints", "url": "https://openrouter.ai/deepseek/deepseek-v4.1-flash-20261017/providers", "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "orq-wafer", "what": "the quantization OpenRouter lists for Wafer's DeepSeek V4.1 Flash 1017 endpoint", "value": "mxfp4", "shown": "mxfp4", "whose": "OpenRouter", "source": "OpenRouter, DeepSeek V4.1 Flash 1017 endpoints", "url": "https://openrouter.ai/deepseek/deepseek-v4.1-flash-20261017/providers", "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-tot-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: output tokens over all six runs", "value": 257813, "shown": "257,813", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-pagoda-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: output tokens over the two pagoda runs", "value": 89242, "shown": "89,242", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-platformer-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: output tokens over the two platformer runs", "value": 90757, "shown": "90,757", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-solar-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: output tokens over the two solar runs", "value": 77814, "shown": "77,814", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-usd-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: the six runs priced at list prices (our arithmetic; not what we paid)", "value": 0.15476625, "shown": "$0.155", "whose": "Model Fatigue, arithmetic on our runs", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-in-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: input tokens over all six runs, cached included", "value": 1576, "shown": "1,576", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-tot-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: output tokens over all six runs", "value": 145487, "shown": "145,487", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-pagoda-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: output tokens over the two pagoda runs", "value": 49065, "shown": "49,065", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-platformer-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: output tokens over the two platformer runs", "value": 65444, "shown": "65,444", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-solar-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: output tokens over the two solar runs", "value": 30978, "shown": "30,978", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-usd-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: the six runs priced at list prices (DeepSeek off-peak) (our arithmetic; not what we paid)", "value": 0.0873669, "shown": "$0.087", "whose": "Model Fatigue, arithmetic on our runs", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-in-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: input tokens over all six runs, cached included", "value": 498, "shown": "498", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-demo", "what": "platformer: when each of the four demos started (no key pressed)", "value": 3.0, "shown": "3.0 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-ds1-over", "what": "platformer, DeepSeek run 1: when the demo ran into the enemy", "value": "8.0–8.2 s", "shown": "8.0–8.2 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-ds2-over", "what": "platformer, DeepSeek run 2: when the demo ran into the enemy", "value": "8.4–8.6 s", "shown": "8.4–8.6 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-ds1-re", "what": "platformer, DeepSeek run 1: when the level restarted", "value": 10.0, "shown": "10.0 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-ds2-re", "what": "platformer, DeepSeek run 2: when the level restarted", "value": 10.5, "shown": "10.5 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-rec", "what": "length of each recording", "value": 14, "shown": "14 s", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-mm1-score", "what": "platformer, MiniMax run 1: score at the end of the recording", "value": "SCORE 2000, COINS 20/41", "shown": "SCORE 2000, COINS 20/41", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-mm2-score", "what": "platformer, MiniMax run 2: score at the end of the recording", "value": "SCORE 2300, COINS 23/40", "shown": "SCORE 2300, COINS 23/40", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "pf-mm-delay", "what": "platformer, MiniMax run 1: the idle time before its demo, in milliseconds", "value": 3000, "shown": "3,000", "whose": "Model Fatigue (our test)", "source": "Our checklist audit of the side-by-side pages (each page against each ask in its prompt, from the recording and the code)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "per-worst", "what": "the largest difference from NASA's period on any planet of any of the four pages", "value": 0.006437458193979895, "shown": "0.64%", "whose": "Model Fatigue, arithmetic on the pages and NASA's sheet", "source": "Our check of the solar pages' orbital periods against NASA's planetary fact sheet", "url": "https://nssdc.gsfc.nasa.gov/planetary/factsheet/", "read": "2026-10-20T04:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-sun", "what": "the Sun's geocentric ecliptic longitude at 2026-10-20 00:00 UTC (JPL DE440s), degrees", "value": 206.618, "shown": "206.6", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-moon", "what": "the Moon's geocentric ecliptic longitude at that moment, degrees", "value": 311.198, "shown": "311.2", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-lit", "what": "the Moon's illuminated fraction at that moment", "value": 0.6271, "shown": "0.627", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-mm2-label", "what": "the Moon's longitude MiniMax wrote on its label in run 2", "value": "311.28°", "shown": "311.28°", "whose": "MiniMax M3.1 Flash (its output in our test)", "source": "MiniMax M3.1 Flash's answer to our Moon prompt, run 2 (the SVG as it came back)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-mm2-drawn", "what": "where MiniMax's run 2 drew the Moon, degrees of longitude (our reading of its SVG)", "value": 48.72810063642653, "shown": "48.7", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-cradle-mm", "what": "Newton's cradle, MiniMax: presses right over three runs, of nine", "value": 6, "shown": "6", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-cradle-ds", "what": "Newton's cradle, DeepSeek: presses right over three runs, of nine", "value": 9, "shown": "9", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-cradle-line", "what": "the line in MiniMax's run 2 cradle that takes one off the key pressed", "value": "launch(Number(event.key) - 1);", "shown": "launch(Number(event.key) - 1);", "whose": "MiniMax M3.1 Flash (its output in our test)", "source": "MiniMax M3.1 Flash's answer to our cradle prompt, run 2 (line 482 of the page it wrote)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-cradle-keys", "what": "the keys our cradle prompt asks the page to respond to", "value": "1, 2 or 3", "shown": "1, 2 or 3", "whose": "Model Fatigue (our test)", "source": "Our Newton's cradle prompt", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-gear-teeth", "what": "the teeth on the five gears our prompt asks for, left to right", "value": "12, 30, 18, 40 and 24", "shown": "12, 30, 18, 40 and 24", "whose": "Model Fatigue (our test)", "source": "The prompts of our tests with a right answer", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-gear-rpm", "what": "the driven gear's speed in our gears prompt, rpm", "value": 60, "shown": "60", "whose": "Model Fatigue (our test)", "source": "The prompts of our tests with a right answer", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-pl-sun", "what": "teeth on the sun gear in our planetary prompt", "value": 24, "shown": "24", "whose": "Model Fatigue (our test)", "source": "Our planetary gearbox prompt", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-pl-planet", "what": "teeth on each planet gear", "value": 18, "shown": "18", "whose": "Model Fatigue (our test)", "source": "Our planetary gearbox prompt", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-pl-ring", "what": "teeth on the ring gear", "value": 60, "shown": "60", "whose": "Model Fatigue (our test)", "source": "Our planetary gearbox prompt", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-bhavani-views", "what": "views of a whale breaching post (2112423796527559136) at model-economics' pull of 21 October", "value": 45933, "shown": "45,933", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rubens-views", "what": "views of a campfire post (2112211798493397408) at model-economics' pull of 21 October", "value": 44043, "shown": "44,043", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-atomic-views", "what": "views of a police chase post (2112263410301964712) at model-economics' pull of 21 October", "value": 42284, "shown": "42,284", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-roadster-views", "what": "views of a roadster drive post (2112247645955895455) at model-economics' pull of 21 October", "value": 21602, "shown": "21,602", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-eiffel-views", "what": "views of the Eiffel Tower and Paris post (2112304943076184535) at model-economics' pull of 21 October", "value": 15524, "shown": "15,524", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-fawn-views", "what": "views of a deer post (2112217467775062514) at model-economics' pull of 21 October", "value": 8407, "shown": "8,407", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-rocket-views", "what": "views of a rocket launch post (2112573104996089906) at model-economics' pull of 21 October", "value": 5833, "shown": "5,833", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-alpine-views", "what": "views of an alpine flight post (2112624894680645801) at model-economics' pull of 21 October", "value": 5519, "shown": "5,519", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-bhavani-mm", "what": "MiniMax M3.1 Flash's cost in the post on a whale breaching", "value": 0.25, "shown": "$0.25", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-bhavani-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a whale breaching", "value": 0.76, "shown": "$0.76", "whose": "the posters on X (each credited by handle)", "source": "Costs shown only in two posts' own videos (@Bhavani_00007's whale, @EnvolDev's alpine flight), read from a frame", "url": null, "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rubens-mm", "what": "MiniMax M3.1 Flash's cost in the post on a campfire", "value": 0.22, "shown": "$0.22", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rubens-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a campfire", "value": 0.54, "shown": "$0.54", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-atomic-mm", "what": "MiniMax M3.1 Flash's cost in the post on a police chase", "value": 2.34, "shown": "$2.34", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-atomic-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a police chase", "value": 3.07, "shown": "$3.07", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-roadster-mm", "what": "MiniMax M3.1 Flash's cost in the post on a roadster drive (“~”, about, as written)", "value": 1.37, "shown": "$1.37", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-roadster-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a roadster drive", "value": 0.66, "shown": "$0.66", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-eiffel-mm", "what": "MiniMax M3.1 Flash's cost in the post on the Eiffel Tower and Paris", "value": 0.24, "shown": "$0.24", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-eiffel-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on the Eiffel Tower and Paris", "value": 0.53, "shown": "$0.53", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-fawn-mm", "what": "MiniMax M3.1 Flash's cost in the post on a deer", "value": 0.029, "shown": "$0.029", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-fawn-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a deer", "value": 0.067, "shown": "$0.067", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-rocket-mm", "what": "MiniMax M3.1 Flash's cost in the post on a rocket launch (“~”, about, as written)", "value": 1.2, "shown": "$1.20", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rafa-rocket-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on a rocket launch", "value": 0.86, "shown": "$0.86", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-alpine-mm", "what": "MiniMax M3.1 Flash's cost in the post on an alpine flight", "value": 0.23, "shown": "$0.23", "whose": "the posters on X (each credited by handle)", "source": "Costs shown only in two posts' own videos (@Bhavani_00007's whale, @EnvolDev's alpine flight), read from a frame", "url": null, "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-envol-alpine-ds", "what": "DeepSeek V4.1 Flash 1017's cost in the post on an alpine flight", "value": 0.49, "shown": "$0.49", "whose": "the posters on X (each credited by handle)", "source": "Costs shown only in two posts' own videos (@Bhavani_00007's whale, @EnvolDev's alpine flight), read from a frame", "url": null, "read": "2026-10-25", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-n", "what": "posts in model-economics' X pull of 21 October (searches, not all of X)", "value": 62, "shown": "62", "whose": "the posters on X (each credited by handle)", "source": "Posts on X comparing the two models, from our searches of X on 21 October (each post linked in the text)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-bhavani-q", "what": "@Bhavani_00007 on MiniMax's whale (four models in the post; the saved JSON escapes the apostrophe in the rest of the sentence)", "value": "the result looks better than the others", "shown": "the result looks better than the others", "whose": "the posters on X (each credited by handle)", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-notjazii-q2", "what": "@notjazii on M3.1 Flash against M3", "value": "a huge step up from m3", "shown": "a huge step up from m3", "whose": "@notjazii", "source": "J A Z I I (@notjazii) on X, 19 October 2026, “first impression of minimax m3.1 flash”", "url": "https://x.com/notjazii/status/2112200201507311699", "read": "2026-10-19T15:50", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-notjazii-q3", "what": "@notjazii on one test against DeepSeek", "value": "in this test it did better than deepseek on first try", "shown": "in this test it did better than deepseek on first try", "whose": "@notjazii", "source": "J A Z I I (@notjazii) on X, 19 October 2026, “first impression of minimax m3.1 flash”", "url": "https://x.com/notjazii/status/2112200201507311699", "read": "2026-10-19T15:50", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-notjazii-q", "what": "@notjazii after a few hours with M3.1 Flash", "value": "hallucinates alot on xhigh or max and starts doing stuff i didn't ask for", "shown": "hallucinates alot on xhigh or max and starts doing stuff i didn't ask for", "whose": "@notjazii", "source": "J A Z I I (@notjazii) on X, 19 October 2026, “first impression of minimax m3.1 flash”", "url": "https://x.com/notjazii/status/2112200201507311699", "read": "2026-10-19T15:50", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-ds-rank", "what": "LMArena WebDev, DeepSeek V4.1 Flash 1017 (Max): rank", "value": 14, "shown": "14", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-ds-score", "what": "LMArena WebDev, DeepSeek V4.1 Flash 1017 (Max): score", "value": 1665, "shown": "1665", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-ds-ci", "what": "LMArena WebDev, DeepSeek V4.1 Flash 1017 (Max): the score's interval, plus or minus", "value": 13, "shown": "13", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-ds-votes", "what": "LMArena WebDev, DeepSeek V4.1 Flash 1017 (Max): votes", "value": 2500, "shown": "2,500", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-mm-rank", "what": "LMArena WebDev, MiniMax M3.1 Flash (preliminary): rank", "value": 34, "shown": "34", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-mm-score", "what": "LMArena WebDev, MiniMax M3.1 Flash (preliminary): score", "value": 1589, "shown": "1589", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-mm-ci", "what": "LMArena WebDev, MiniMax M3.1 Flash (preliminary): the score's interval, plus or minus", "value": 13, "shown": "13", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-mm-votes", "what": "LMArena WebDev, MiniMax M3.1 Flash (preliminary): votes", "value": 2212, "shown": "2,212", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-hk-rank", "what": "LMArena WebDev, Claude Haiku 5.5 (High): rank", "value": 35, "shown": "35", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-hk-score", "what": "LMArena WebDev, Claude Haiku 5.5 (High): score", "value": 1586, "shown": "1586", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-hk-ci", "what": "LMArena WebDev, Claude Haiku 5.5 (High): the score's interval, plus or minus", "value": 10, "shown": "10", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "lm-hk-votes", "what": "LMArena WebDev, Claude Haiku 5.5 (High): votes", "value": 3917, "shown": "3,917", "whose": "LMArena", "source": "LMArena, Code Arena WebDev leaderboard (board dated 22 October 2026)", "url": "https://lmarena.ai/leaderboard/webdev", "read": "2026-10-25T02:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-peak-h", "what": "hours of DeepSeek's peak rate on a weekday (01:00–04:00 and 06:00–10:00 UTC) (our arithmetic)", "value": 7, "shown": "7", "whose": "Model Fatigue, arithmetic on DeepSeek's price page", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-cache", "what": "MiniMax's cached-input price against DeepSeek's off-peak one (our arithmetic)", "value": 5.0, "shown": "5×", "whose": "Model Fatigue, arithmetic on the two price pages", "source": "DeepSeek, Models & Pricing", "url": "https://api-docs.deepseek.com/quick_start/pricing", "read": "2026-10-17T23:06", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-m3", "what": "M3.1 Flash's price per token as a share of M3's (our arithmetic)", "value": 0.5, "shown": "50%", "whose": "Model Fatigue, arithmetic on MiniMax's price page", "source": "MiniMax, API pricing overview", "url": "https://platform.minimax.io/docs/pricing/overview", "read": "2026-10-19T17:10", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "w-mm", "what": "AA-Omniscience, MiniMax M3.1 Flash: share of all questions answered wrong (our arithmetic)", "value": 0.4065, "shown": "40.6%", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "w-ds", "what": "AA-Omniscience, DeepSeek V4.1 Flash 1017: share of all questions answered wrong (our arithmetic)", "value": 0.3815, "shown": "38.1%", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cost-off", "what": "DeepSeek V4.1 Flash 1017's cost per index task at its off-peak rates (half of AA's figure) (our arithmetic)", "value": 0.1310242136557086, "shown": "$0.131", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures and DeepSeek's price page", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cin-off", "what": "the input part of it, off-peak (our arithmetic)", "value": 0.07732718231817783, "shown": "$0.077", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures and DeepSeek's price page", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cr-off", "what": "the cache-read part of it, off-peak (our arithmetic)", "value": 0.027958591958119644, "shown": "$0.028", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures and DeepSeek's price page", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "ds-cout-off", "what": "the output part of it, off-peak (our arithmetic)", "value": 0.05369703133753079, "shown": "$0.054", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures and DeepSeek's price page", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-out-aa", "what": "output tokens over AA's index run, DeepSeek against MiniMax (our arithmetic)", "value": 1.1549747508797599, "shown": "1.15×", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-m3-cost", "what": "how much lower M3.1 Flash's cost per index task is than M3's on Artificial Analysis (our arithmetic)", "value": 0.5663749572113319, "shown": "57%", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-spd", "what": "median output speed on 25 October, DeepSeek against MiniMax (our arithmetic)", "value": 2.0423926951871643, "shown": "2.0×", "whose": "Model Fatigue, arithmetic on Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": true, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-auto-d", "what": "MiniMax's table, Automation-Bench v1.0.6: M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": 2.299999999999997, "shown": "+2.3 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-tb-d", "what": "MiniMax's table, Terminal Bench 4.0: M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": -4.199999999999996, "shown": "−4.2 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-deepswe-d", "what": "MiniMax's table, DeepSWE v1.1: M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": -3.299999999999997, "shown": "−3.3 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-nl2repo-d", "what": "MiniMax's table, NL2Repo-Bench: M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": -7.0, "shown": "−7.0 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-babyvision-d", "what": "MiniMax's table, BabyVision (w/tools): M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": -1.5, "shown": "−1.5 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "mmt-hle-d", "what": "MiniMax's table, HLE (w/o tools): M3.1 Flash minus DeepSeek 1017, points (our arithmetic)", "value": 0.7999999999999972, "shown": "+0.8 points", "whose": "Model Fatigue, arithmetic on MiniMax's table", "source": "MiniMax, its own benchmark table as the Hugging Face model card of 23 October gives it, M3.1 Flash and DeepSeek 1017 columns", "url": "https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "tb-gap-mm", "what": "Terminal-Bench 4.0, MiniMax's own run minus Artificial Analysis's, points (our arithmetic)", "value": 12.342424242424201, "shown": "+12.3 points", "whose": "Model Fatigue, arithmetic on the companies' and Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "tbs-mm", "what": "Terminal-Bench 4.0 on Artificial Analysis, MiniMax M3.1 Flash, as a score out of 100 (our arithmetic)", "value": 25.7575757575758, "shown": "25.8", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "tb-gap-ds", "what": "Terminal-Bench 4.0, DeepSeek's own run minus Artificial Analysis's, points (our arithmetic)", "value": 8.461616161616199, "shown": "+8.5 points", "whose": "Model Fatigue, arithmetic on the companies' and Artificial Analysis's figures", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "tbs-ds", "what": "Terminal-Bench 4.0 on Artificial Analysis, DeepSeek V4.1 Flash 1017, as a score out of 100 (our arithmetic)", "value": 33.8383838383838, "shown": "33.8", "whose": "Artificial Analysis", "source": "Artificial Analysis, MiniMax M3.1 Flash model page (with DeepSeek V4.1 Flash 1017 and MiniMax M3), read 25 October", "url": "https://artificialanalysis.ai/models/minimax-m3-1-flash", "read": "2026-10-25T01:57", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-wmin-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: the quickest run, minutes (our arithmetic)", "value": 5.08639303445816, "shown": "5.1", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-wmax-mm", "what": "our side-by-side test, MiniMax M3.1 Flash: the slowest run, minutes (our arithmetic)", "value": 8.147210200627645, "shown": "8.1", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-wmin-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: the quickest run, minutes (our arithmetic)", "value": 1.2176287651062012, "shown": "1.2", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "sbs-wmax-ds", "what": "our side-by-side test, DeepSeek V4.1 Flash 1017: the slowest run, minutes (our arithmetic)", "value": 4.431918183962504, "shown": "4.4", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-sbs", "what": "our side-by-side test: MiniMax's output tokens against DeepSeek's, all six runs (our arithmetic)", "value": 1.772068982108367, "shown": "1.77×", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-sbs-pagoda", "what": "our side-by-side test, pagoda: MiniMax's output tokens against DeepSeek's (our arithmetic)", "value": 1.8188525425456028, "shown": "1.8×", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-sbs-platformer", "what": "our side-by-side test, platformer: MiniMax's output tokens against DeepSeek's (our arithmetic)", "value": 1.386788704846892, "shown": "1.4×", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-sbs-solar", "what": "our side-by-side test, solar: MiniMax's output tokens against DeepSeek's (our arithmetic)", "value": 2.5119116792562464, "shown": "2.5×", "whose": "Model Fatigue (our test)", "source": "Our side-by-side test: three prompts, two runs per model, default settings (19 October)", "url": null, "read": "2026-10-19", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-mirror", "what": "the Moon's true longitude mirrored across the 0°–180° line, degrees (our arithmetic)", "value": 48.80200000000002, "shown": "48.8", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-min-mm", "what": "the Moon, MiniMax: the quickest run, minutes (our arithmetic)", "value": 7.33, "shown": "7.3", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-max-mm", "what": "the Moon, MiniMax: the slowest run, minutes (our arithmetic)", "value": 18.983333333333334, "shown": "19.0", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-min-ds", "what": "the Moon, DeepSeek: the quickest run, minutes (our arithmetic)", "value": 2.3699999999999997, "shown": "2.4", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-moon-max-ds", "what": "the Moon, DeepSeek: the slowest run, minutes (our arithmetic)", "value": 3.0616666666666665, "shown": "3.1", "whose": "Model Fatigue (our test)", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "vt-pl-carrier", "what": "the carrier's true speed with the ring held and the sun at 60 rpm, rpm (our arithmetic)", "value": 17.142857142857142, "shown": "17.14", "whose": "Model Fatigue, arithmetic on the prompt's gears", "source": "Our tests with a right answer: gears, a planetary gearbox, a Newton's cradle and the Moon, three runs per model (20 October)", "url": null, "read": "2026-10-20", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-bhavani-r", "what": "MiniMax's cost as a share of DeepSeek's in the post on a whale breaching (our arithmetic)", "value": 0.32894736842105265, "shown": "33%", "whose": "Model Fatigue, arithmetic on the posters' figures", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-rubens-r", "what": "MiniMax's cost as a share of DeepSeek's in the post on a campfire (our arithmetic)", "value": 0.4074074074074074, "shown": "41%", "whose": "Model Fatigue, arithmetic on the posters' figures", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null }, { "id": "x-atomic-r", "what": "MiniMax's cost as a share of DeepSeek's in the post on a police chase (our arithmetic)", "value": 0.762214983713355, "shown": "76%", "whose": "Model Fatigue, arithmetic on the posters' figures", "source": "The posts' own words (the same X pull of 21 October)", "url": null, "read": "2026-10-21T09:18", "moves": false, "caveat": null, "now": null, "now_shown": null, "now_read": null } ] }