# Vals AI Access to Justice (Access to Justice)

> Can AI help someone navigate a civil legal problem without a lawyer?

Canonical page: https://benchlm.ai/benchmarks/access-to-justice

- Category: [Agentic](/agentic)
- Last updated: November 10, 2026

Evaluation results published by [Vals AI](https://www.vals.ai/benchmarks/a2j). Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.

## About Access to Justice

- Year: 2026
- Tasks: 252 simulated legal-aid conversations across 8 topics
- Format: Share of expert rubric criteria met
- Difficulty: First-contact civil legal help with local referrals
- Paper: [Access to Justice leaderboard and methodology](https://www.vals.ai/benchmarks/a2j)

Vals published Access to Justice on November 10, 2026. Each of 252 scenarios starts from a real query brought to a legal aid organization and runs as a simulated conversation; legal aid experts wrote the grading criteria. A good first response asks about the person's situation, gives accurate general legal information, and steers the person to the right local help such as a legal aid program or a court self-help center. The score is the share of rubric criteria met. BenchLM mirrors the published aggregates as display-only external evidence, outside overall and category rankings.

Access to Justice is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (23 models)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [Qwen3.8 Max](/models/qwen3-8-max) | — | Alibaba | 76.37% |
| 2 | [Claude Sonnet 5.5](/models/claude-sonnet-5-5) | — | Anthropic | 74.94% |
| 3 | [Muse Wonder Full](/models/muse-wonder-full) | max reasoning | Meta | 74.68% |
| 4 | [Claude Opus 5.5](/models/claude-opus-5-5) | — | Anthropic | 73.77% |
| 5 | [GLM-5.3](/models/glm-5-3) | max reasoning | Z.AI | 73.70% |
| 6 | [Grok 4.7](/models/grok-4-7) | xhigh reasoning | xAI | 73.69% |
| 7 | [MiniMax M3.1 Flash](/models/minimax-m3-1-flash) | — | MiniMax | 73.31% |
| 8 | [DeepSeek V4.1 Flash 1017](/models/deepseek-v4-1-flash) | max reasoning | DeepSeek | 73.11% |
| 9 | [Muse Wonder](/models/muse-wonder) | max reasoning | Meta | 72.96% |
| 10 | [DeepSeek V4.1 Flash 1017](/models/deepseek-v4-1-flash) | max reasoning | DeepSeek | 72.86% |
| 11 | [Muse Spark 1.3](/models/muse-spark-1-3) | xhigh reasoning | Meta | 72.18% |
| 12 | [Muse Spark 1.3 Max](https://www.vals.ai/models/meta_muse_spark_1_3_max) | max reasoning | Meta | 72.15% |
| 13 | [Grok 4.6](/models/grok-4-6) | high reasoning | xAI | 71.34% |
| 14 | [GLM-5.3-Flash](/models/glm-5-3-flash) | max reasoning | Z.AI | 71.11% |
| 15 | [MiMo-V2.6-Flash](/models/mimo-v2-6-flash) | — | Xiaomi | 70.75% |
| 16 | [MiMo-V2.6-Pro](/models/mimo-v2-6-pro) | — | Xiaomi | 69.96% |
| 17 | [MiniMax M3](/models/minimax-m3) | — | MiniMax | 69.77% |
| 18 | [Kimi K3.1 Pro](/models/kimi-3-1-pro) | max reasoning | Moonshot AI | 69.47% |
| 19 | [Gemini 4 Argon](/models/gemini-4-argon) | high reasoning | Google | 68.48% |
| 20 | [Kimi K3](/models/kimi-k3) | max reasoning | Moonshot AI | 68.17% |
| 21 | [Gemini 3.8 Flash](/models/gemini-3-8-flash) | high reasoning | Google | 67.06% |
| 22 | [GPT-6.1 Sol](/models/gpt-6-1-sol) | max reasoning | OpenAI | 65.84% |
| 23 | [GPT-5.6 Luna](/models/gpt-5-6-luna) | max reasoning | OpenAI | 65.56% |

## FAQ

### What does Access to Justice measure?

Can AI help someone navigate a civil legal problem without a lawyer?

### Which model leads the published Access to Justice snapshot?

Qwen3.8 Max currently leads the published Access to Justice snapshot with a score of 76.37%.

### How many models are evaluated on Access to Justice?

The November 10, 2026 contains 23 AI models.

### Does Access to Justice affect BenchLM's overall score?

Not directly. Access to Justice is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.