Files
agent-runtime-framework/benchmarks/artificial-analysis/2026-10-13.csv
Paul O'Reilly 418d634afa benchmarks: add Artificial Analysis snapshot and placeholder M3.1-Flash estimate
- artificial-analysis/2026-10-13.csv: Anthropic, Qwen, MiniMax-M3 rows with
  cost per task and derived cost per intelligence point
- minimax-m3-1-flash-unverified.md: operator-supplied KingBench figures, unverified
- estimates.csv: TODO placeholder for M3.1 Flash (0.9x M3 cost, index scaled by
  KingBench ratio) -- replace with measured/published values
2026-10-13 14:33:53 +13:00

2.1 KiB

1modelcreatorcontextintelligence_indexcost_per_task_usdcost_per_intelligence_point_usdmedian_tokens_per_s
2Claude Opus 5.5 (max with fallback)Anthropic1M585.980.103196
3Claude Sonnet 5.5 (max with fallback)Anthropic1M565.460.0975135
4Claude Opus 5.5 (xhigh with fallback)Anthropic1M563.460.061881
5Claude Opus 5.5 (high with fallback)Anthropic1M541.820.033782
6Claude Fable 5.1 (max with fallback)Anthropic1M537.630.144067
7Claude Fable 5.1 (xhigh with fallback)Anthropic1M535.980.112862
8Claude Sonnet 5.5 (xhigh with fallback)Anthropic1M522.010.0387103
9Claude Opus 5.5 (medium with fallback)Anthropic1M511.340.026379
10Claude Fable 5.1 (high with fallback)Anthropic1M513.910.076752
11Claude Fable 5.1 (medium with fallback)Anthropic1M492.980.060854
12Claude Fable 5.1 (low with fallback)Anthropic1M472.370.050456
13Claude Sonnet 5.5 (high with fallback)Anthropic1M470.880.018796
14Claude Haiku 5.5 (max)Anthropic1M430.210.0049236
15Claude Opus 5.5 (low with fallback)Anthropic1M420.550.013180
16Claude Haiku 5.5 (xhigh)Anthropic1M410.120.0029187
17Claude Sonnet 5.5 (medium with fallback)Anthropic1M410.480.011795
18Claude Haiku 5.5 (high)Anthropic1M380.080.0021175
19Claude Sonnet 5.5 (low with fallback)Anthropic1M360.350.0097100
20Claude Haiku 5.5 (medium)Anthropic1M340.050.0015171
21Claude Haiku 5.5 (low)Anthropic1M290.020.0007164
22MiniMax-M3MiniMax1M290.510.0176107
23Qwen3.8 Max (0902)Alibaba984k455.410.120236
24Qwen3.8 2.4T A95BAlibaba262k402.160.054037
25Qwen3.8-Flash-NextAlibaba256k400.370.009251
26Qwen3.8 27B (xhigh)Alibaba256k341.010.029746
27Qwen3.8 27B (medium)Alibaba256k281.130.040450
28Qwen3.8 27B (low)Alibaba256k261.050.040453
29Qwen3.7 PlusAlibaba1M250.220.008855
30Qwen3.8 27B (non-reasoning)Alibaba256k202.490.124552
31Qwen3.5 397B A17BAlibaba262k180.470.026188
32Qwen3.6 35B A3BAlibaba262k180.480.0267129
33Qwen3.5 122B A10BAlibaba262k160.320.0200129
34Qwen3.5 9BAlibaba262k110.210.019158
35Qwen3 Coder NextAlibaba256k90.780.086773