HELM Lite - LegalBench - PROA: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 91 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.1 70B Instruct | 100 |
| 2 | Mistral Large 2 (Jul) | 100 |
| 3 | Gemini 1.5 Pro (001) | 100 |
| 4 | Claude 3.5 Sonnet (20241022) | 98.95 |
| 5 | GPT-4o (2024-05-13) | 98.95 |
| 6 | Gemini 2.0 Flash (Preview) | 98.95 |
| 7 | Claude 3 Opus (20240229) | 98.95 |
| 8 | Llama 3.2 90B Vision Instruct | 98.95 |
| 9 | GPT-4 Turbo (Preview) | 98.95 |
| 10 | Jamba 1.5 Large | 98.95 |
| 11 | Llama 3.3 70B Instruct | 97.89 |
| 12 | Qwen 2.5 72B Instruct | 97.89 |
| 13 | Gemini 1.5 Flash (002) | 97.89 |
| 14 | GPT-4o (2024-08-06) | 97.89 |
| 15 | Gemma 2 27B (IT) | 97.89 |
Interactive version: theaggregate.ai/benchmark?slug=helm-lite-legalbench-proa · How It Works · Data refreshed daily, snapshot 2026-09-19.