Benchmark
Vals Access to Justice Benchmark
Scores are percentages.
Benchmark sourceScores from Vals AI.
- Models measured
- 22
- Scores recorded
- 22
- Sources
- 1
- IQ weight
- 0.16
Score distribution
- 71.3%
- Mean
- 72.2%
- Median
- 2.9%
- Std dev
- 67.1%
- 10th percentile
- 74.7%
- 90th percentile
- 65.6%
- Min
This benchmark has a weight of 0.16 for chat models, measured over 22 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Qwen3.8 Max· Vals AI | 76.4% |
| 2 | Claude Sonnet 5.5· Vals AI | 74.9% |
| 3 | Muse Wonder Full· Vals AI | 74.7% |
| 4 | Claude Opus 5.5· Vals AI | 73.8% |
| 5 | GLM 5.3· Vals AI | 73.7% |
| 6 | Grok 4.7· Vals AI | 73.7% |
| 7 | MiniMax M3.1 Flash· Vals AI | 73.3% |
| 8 | DeepSeek V4.1 Flash 1017· Vals AI | 73.1% |
| 9 | Muse Wonder· Vals AI | 73% |
| 10 | DeepSeek V4.1 Flash· Vals AI | 72.9% |
| 11 | Muse Spark 1.3· Vals AI | 72.2% |
| 12 | Grok 4.6· Vals AI | 71.3% |
| 13 | GLM 5.3 Flash· Vals AI | 71.1% |
| 14 | MiMo V2.6 Flash· Vals AI | 70.7% |
| 15 | MiMo V2.6 Pro· Vals AI | 70% |
| 16 | MiniMax M3· Vals AI | 69.8% |
| 17 | Kimi K3.1 Pro· Vals AI | 69.5% |
| 18 | Gemini 4 Argon· Vals AI | 68.5% |
| 19 | Kimi K3· Vals AI | 68.2% |
| 20 | Gemini 3.8 Flash· Vals AI | 67.1% |
| 21 | GPT-6.1 Sol· Vals AI | 65.8% |
| 22 | GPT-5.6 Luna· Vals AI | 65.6% |