Vals AI Immigration: leaderboard
Vals Immigration evaluates U.S. immigration-law work through up to five exchanges with a simulated client. The main no-search result averages rubric credit on 262 private tasks; unfinished tasks score zero.
Metric: Mean Rubric Credit (%). Source: www.vals.ai. Saturation forecast: Around 2030. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5.1 (Max) | 75.44 | #10 (Claude Fable 5.1) |
| 2 | Claude Opus 5.5 (Max) | 72.14 | #3 (Claude Opus 5.5) |
| 3 | Muse Wonder Full (Max) | 71.22 | #11 (Muse Wonder Full) |
| 4 | Qwen 3.8 Max | 70.14 | #33 |
| 5 | Muse Wonder (Max) | 69.62 | #17 (Muse Wonder) |
| 6 | Claude Sonnet 5.5 (Max) | 68.96 | #12 (Claude Sonnet 5.5) |
| 7 | Kimi K3 (Max) | 68.26 | #35 (Kimi K3) |
| 8 | Gemini 4 Argon (High) | 68.08 | #9 (Gemini 4 Argon) |
| 9 | Grok 4.7 (xHigh) | 68.05 | #27 (Grok 4.7) |
| 10 | GPT-6.1 Sol (Max) | 67.81 | #7 (GPT-6.1 Sol) |
| 11 | GPT-6 Astra (Max) | 67.69 | #5 (GPT-6 Astra) |
| 12 | MiMo-V2.6-Pro | 67.5 | #49 |
| 13 | GLM-5.3 (Max) | 66.73 | #42 (GLM-5.3) |
| 14 | Muse Spark 1.2 (xHigh) | 66.54 | #68 (Muse Spark 1.2) |
| 15 | Grok 4.6 (High) | 65.98 | #30 (Grok 4.6) |
Interactive version: theaggregate.ai/benchmark?slug=vals-ai-immigration · How It Works · Data refreshed daily, snapshot 2026-10-29.