French LLM Leaderboard - Average — leaderboard
French LLM leaderboard aggregating IFEval FR, GPQA FR, and BAC FR to compare open models and chatbots on French instruction-following, graduate science QA, and exam-style reasoning.
Metric: Average Score (%). Source: huggingface.co. Status: saturation imminent. 55 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 Distill Llama 70B | 60.02 |
| 2 | Mistral Large 2 (Nov) Instruct (2411) | 55.19 |
| 3 | Llama 3.3 70B Instruct | 54.72 |
| 4 | DeepSeek R1 Distill Qwen 32B | 53.39 |
| 5 | Qwen 2.5 72B Instruct | 53.2 |
| 6 | calme-3.2-instruct-78B | 49.89 |
| 7 | Qwen 2.5 14B Instruct | 48.9 |
| 8 | Chocolatine-2-14B-Instruct-v2.0 | 48.71 |
| 9 | Phi-4 | 48.21 |
| 10 | Virtuoso-Lite | 46.8 |
| 11 | Mistral Small 3 | 45.02 |
| 12 | Mixtral 8x22B Instruct (v0.1) | 44.73 |
| 13 | Phi-3.5-MoE-instruct | 40.16 |
| 14 | Phi-3-medium-128k-instruct | 36.76 |
| 15 | calme-3.3-instruct-3B | 35.59 |
Interactive version: theaggregate.ai/benchmark?slug=french-llm-leaderboard-average · How the rankings work · Data refreshed daily, snapshot 2026-07-22.