Vals AI Legal Index: leaderboard
Vals AI legal capability composite: Legal Research Bench, HLAB Vals v1.1 and Tax Agent Bench, weighted by 2022 US law-office revenue across practice areas. Published cohort and weighted accuracy, not a simple mean of component leaderboard scores.
Metric: Legal Index (%). Source: www.vals.ai. Saturation forecast: Not forecast. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Muse Wonder Full (Max) | 77.86 | #10 (Muse Wonder Full) |
| 2 | Gemini 4 Argon (High) | 76.72 | #8 (Gemini 4 Argon) |
| 3 | Muse Wonder (Max) | 75.26 | #15 (Muse Wonder) |
| 4 | GLM-5.3 (Max) | 74.58 | #38 (GLM-5.3) |
| 5 | Grok 4.7 (xHigh) | 72.6 | #26 (Grok 4.7) |
| 6 | Claude Opus 5.5 (Max) | 72.33 | #5 (Claude Opus 5.5) |
| 7 | MiMo-V2.6-Pro | 72.14 | #44 |
| 8 | DeepSeek V4.1 Flash (1017) (Max) | 71.86 | #25 (DeepSeek V4.1 Flash (1017)) |
| 9 | Kimi K3 (Max) | 71.72 | #33 (Kimi K3) |
| 10 | Hy4 preview | 71.7 | #41 |
| 11 | GLM-5.3 Flash (Max) | 71.05 | #73 (GLM-5.3 Flash) |
| 12 | Claude Haiku 5.5 (Max) | 70.27 | #43 (Claude Haiku 5.5) |
| 13 | Gemini 3.8 Flash (High) | 69.78 | #23 (Gemini 3.8 Flash) |
| 14 | DeepSeek V4 Pro (0813) (Max) | 69.01 | #54 (DeepSeek V4 Pro (0813)) |
| 15 | MiMo-V2.6-Flash | 68.99 | #74 |
No result here: #9 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=vals-ai-legal-index · How It Works · Data refreshed daily, snapshot 2026-10-25.