French LLM Leaderboard - GPQA FR: leaderboard
French GPQA score, measuring graduate-level science question answering in French.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 55 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek R1 Distill Llama 70B | 63.19 | #594 |
| 2 | DeepSeek R1 Distill Qwen 32B | 54.4 | #655 |
| 3 | Mistral Large 2 (Nov) Instruct (2411) | 47.99 | #465 |
| 4 | Phi-4 | 46.52 | #706 |
| 5 | Llama 3.3 70B Instruct | 44.32 | #534 |
| 6 | Qwen 2.5 72B Instruct | 43.59 | #454 |
| 7 | calme-3.2-instruct-78B | 41.03 | #347 |
| 8 | Qwen 2.5 14B Instruct | 36.63 | #647 |
| 9 | Mistral Small 3 | 36.63 | #627 |
| 10 | Mixtral 8x22B Instruct (v0.1) | 36.45 | #756 |
| 11 | Falcon3-10B-Instruct | 33.52 | #1157 |
| 12 | Falcon3-7B-Instruct | 32.78 | #1305 |
| 13 | Phi-3.5-MoE-instruct | 32.6 | #700 |
| 14 | Bielik-11B-v2.3-Instruct | 32.23 | #693 |
| 15 | Phi-3-medium-128k-instruct | 30.4 | #1013 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=french-llm-leaderboard-gpqa-fr · How It Works · Data refreshed daily, snapshot 2026-10-20.