# Vals Immigration Bench (Immigration Bench)

> Vals AI and Talent Mobility Fund benchmark of expert-written U.S. immigration law questions, scored against expert rubrics.

Canonical page: https://benchlm.ai/benchmarks/valsimmigrationbench

- Category: [Knowledge](/knowledge)
- Last updated: October 28, 2026

Evaluation results published by [Vals AI](https://www.vals.ai/benchmarks/immigration). Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.

## About Immigration Bench

- Year: 2026
- Tasks: 262 private test questions
- Format: Average share of expert rubric items met
- Difficulty: Fact-specific U.S. immigration law guidance
- Paper: [Immigration Bench leaderboard and methodology](https://www.vals.ai/benchmarks/immigration)

Vals published Immigration Bench on October 28, 2026. Immigration experts wrote 519 questions from seven personas (international student, scholar or faculty, university advisor, employer, entrepreneur, self-petitioner and fragile or temporary status). Results cover the private 262-question test set; the other 257 questions are reserved for labs to license. Each question has an expert rubric, an AI grader marks each item met when the answer includes it, and the score is the share of rubric items met, averaged over questions. We mirror the published leaderboard as a display-only page, outside overall and category rankings.

Immigration Bench is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (25 models)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [Claude Fable 5.1](/models/claude-fable-5-1) | — | Anthropic | 75.44% |
| 2 | [Claude Opus 5.5](/models/claude-opus-5-5) | — | Anthropic | 72.14% |
| 3 | [Muse Wonder Full](/models/muse-wonder-full) | max reasoning | Meta | 71.22% |
| 4 | [Qwen3.8 Max](/models/qwen3-8-max) | — | Alibaba | 70.14% |
| 5 | [Muse Wonder](/models/muse-wonder) | max reasoning | Meta | 69.62% |
| 6 | [Claude Sonnet 5.5](/models/claude-sonnet-5-5) | — | Anthropic | 68.96% |
| 7 | [Kimi K3](/models/kimi-k3) | max reasoning | Moonshot AI | 68.26% |
| 8 | [Gemini 4 Argon](/models/gemini-4-argon) | high reasoning | Google | 68.08% |
| 9 | [Grok 4.7](/models/grok-4-7) | xhigh reasoning | xAI | 68.05% |
| 10 | [GPT-6.1 Sol](/models/gpt-6-1-sol) | max reasoning | OpenAI | 67.81% |
| 11 | [GPT-6 Astra](/models/gpt-6-astra) | max reasoning | OpenAI | 67.69% |
| 12 | [MiMo-V2.6-Pro](/models/mimo-v2-6-pro) | — | Xiaomi | 67.50% |
| 13 | [GLM-5.3](/models/glm-5-3) | max reasoning | Z.AI | 66.73% |
| 14 | [Muse Spark 1.2](/models/muse-spark-1-2) | xhigh reasoning | Meta | 66.54% |
| 15 | [Grok 4.6](/models/grok-4-6) | high reasoning | xAI | 65.98% |
| 16 | [DeepSeek V4.1 Flash 1017](/models/deepseek-v4-1-flash) | max reasoning | DeepSeek | 65.64% |
| 17 | [Muse Spark 1.3](/models/muse-spark-1-3) | xhigh reasoning | Meta | 65.00% |
| 18 | [Muse Spark 1.3 Max](https://www.vals.ai/models/meta_muse_spark_1_3_max) | max reasoning | Meta | 64.93% |
| 19 | [Mistral Large 4](/models/mistral-large-4) | high reasoning | Mistral | 63.71% |
| 20 | [Gemini 3.8 Flash](/models/gemini-3-8-flash) | high reasoning | Google | 63.41% |
| 21 | [GPT-6 Luna](/models/gpt-6-luna) | max reasoning | OpenAI | 62.53% |
| 22 | [Claude Haiku 5.5](/models/claude-haiku-5-5) | — | Anthropic | 61.79% |
| 23 | [DeepSeek V4 Pro 0813](/models/deepseek-v4-pro-0813) | max reasoning | DeepSeek | 61.66% |
| 24 | [MiniMax M3](/models/minimax-m3) | — | MiniMax | 57.53% |
| 25 | [Gemini 3.5 Flash-Lite](/models/gemini-3-5-flash-lite) | high reasoning | Google | 55.66% |

## FAQ

### What does Immigration Bench measure?

Vals AI and Talent Mobility Fund benchmark of expert-written U.S. immigration law questions, scored against expert rubrics.

### Which model leads the published Immigration Bench snapshot?

Claude Fable 5.1 currently leads the published Immigration Bench snapshot with a score of 75.44%.

### How many models are evaluated on Immigration Bench?

The October 28, 2026 contains 25 AI models.

### Does Immigration Bench affect BenchLM's overall score?

Not directly. Immigration Bench is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.