EuroEval French NLU: leaderboard
French-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: NLU Average Score (%). Source: euroeval.com. Saturation forecast: Around April 2027. 381 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B (IT) | 79.51 | #176 |
| 2 | Gemini 3 Pro (Preview) | 79.33 | #66 |
| 3 | Claude Sonnet 4.6 | 77.21 | #90 |
| 4 | Qwen 3.5 35B A3B Base | 77.13 | #310 |
| 5 | Gemma 4 26B A4B (IT) | 77.12 | #297 |
| 6 | Llama 3.1 70B | 76.9 | #592 |
| 7 | Ministral-3-14B-Reasoning-2512 | 76.61 | #540 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 75.8 | #194 |
| 9 | Gemini 3 Flash (Preview) (Non-reasoning) | 75.55 | #82 (Gemini 3 Flash (Preview)) |
| 10 | Gemini 3 Flash (Preview) (Thinking) | 75.53 | #82 (Gemini 3 Flash (Preview)) |
| 11 | GPT-5.2 | 74.33 | #110 |
| 12 | GPT-5.4 Mini (2026-03-17) (High) | 74.03 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 13 | Ministral-3-14B-Instruct-2512 | 73.63 | #615 |
| 14 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 73.48 | #679 |
| 15 | Grok 4.1 Fast (Reasoning) | 72.72 | #217 (Grok 4.1 Fast) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-french-nlu · How It Works · Data refreshed daily, snapshot 2026-10-20.