Vals Index
Results
26 of 41 tracked models have a result.
- 1Gemini 4 ArgonLimited
Google DeepMind · High effort · $15.7 per test
68.9% - 2
Anthropic · Max effort · $21.3 per test
67.0% - 3
Anthropic · Max effort · $32.1 per test
67.0% - 4
Anthropic · Max effort · $28.7 per test
65.8% - 5
OpenAI · Max effort · $18.5 per test
63.1% - 6
OpenAI · Max effort · $3.24 per test
61.2% - 7
Meta · Max effort · $3.79 per test
58.2% - 8MiMo-V2.6-ProOpen
Xiaomi · $0.41 per test
55.2% - 9
SpaceXAI · xHigh effort · $12.1 per test
55.0% - 10
Google DeepMind · High effort · $5.73 per test
54.8% - 11
Anthropic · Max effort · $2.99 per test
54.3% - 12GLM-5.3Open
Z.ai · Max effort · $7.25 per test
53.5% - 13MiMo-V2.6-FlashOpen
Xiaomi · $0.20 per test
53.2% - 14
OpenAI · Max effort · $5.78 per test
53.1% - 15
DeepSeek · High effort · $0.33 per test
51.3% - 16
OpenAI · Max effort · $0.43 per test
51.2% - 17Kimi K3Open
Moonshot AI · Max effort · $6.38 per test
50.3% - 18Hy4 previewOpen
Tencent · $1.44 per test
49.9% - 19
StepFun · $2.63 per test
49.4% - 20
Alibaba · $4.96 per test
48.3% - 21
Mistral AI · High effort · $13.8 per test
48.1% - 22DeepSeek V4 ProOpen
DeepSeek · Max effort · $3.35 per test
47.6% - 23MiniMax-M3Open
MiniMax · $3.08 per test
36.5% - 24
Google DeepMind · High effort · $2.18 per test
33.4% - 25InklingOpen
Thinking Machines Lab · $1.46 per test
28.7% - 26Inkling-SmallOpen
Thinking Machines Lab · $0.28 per test
25.5%
15 tracked models without a result
Command A+, Gemini 3.5 Flash-Lite, Gemma 4 31B, GLM-5.3-Flash, gpt-oss-120b, Hy3, Kimi K2.7 Code, MAI-Thinking-1, Mistral Medium 3.5, Mistral Small 4, Muse Glimmer 30B, Nemotron 3 Ultra, Nova 2 Lite, Qwen3.8-27B, Qwen3.8-Flash.
Missing means the evaluator has not published a run, not that the model would score poorly.