MGSM French
Math · 2022-10-06
French language split of the MGSM multilingual grade-school math benchmark.
Top models (higher is better)
| Model | Score |
|---|---|
| Opus 4.1 | 93.2 |
| Opus 4 | 92.8 |
| Claude 3.5 Sonnet (Oct 2024) | 92.0 |
| DeepSeek-V3 | 91.6 |
| Llama 4 Maverick Instruct | 91.6 |
| Sonnet 4 | 91.2 |
| DeepSeek-R1 | 90.8 |
| Grok 3 | 90.8 |
| Llama 3.3 70B Instruct | 90.8 |
| DeepSeek-V3-0324 | 90.4 |
| DeepSeek-V3.2 | 90.4 |
| Grok 4 | 90.4 |
| Opus 4.5 | 90.0 |
| O4 Mini | 90.0 |
| Sonnet 4.5 | 89.6 |
| Kimi K2 Instruct | 88.8 |
| Claude 3.7 Sonnet | 88.4 |
| Mistral Large 3 675B Instruct 2512 | 88.0 |
| Qwen3 Max Preview | 88.0 |
| Qwen3-235B-A22B | 87.6 |
| Gemini 3 Flash Preview | 87.2 |
| Llama 4 Scout Instruct | 87.2 |
| GPT-4o Mini | 86.8 |
| GPT-4.1 Mini | 86.4 |
| Grok 3 Mini | 86.0 |
| Qwen3 Max (rolling alias) | 86.0 |
| Command A | 85.6 |
| Gemini 2.5 Flash-Lite Preview (09-2025) | 85.6 |
| GPT-5 Nano | 85.2 |
| GLM-4.5 | 84.8 |
| Grok 2 | 84.8 |
| gpt-oss-20b | 84.4 |
| o3 | 84.4 |
| Haiku 3.5 | 84.0 |
| Haiku 4.5 | 84.0 |
| GPT-4o (2024-08-06) | 84.0 |
| Mistral Large 2.1 (Instruct 2411) | 84.0 |
| O1 | 84.0 |
| Grok 4 Fast | 83.6 |
| MiniMax M2.1 | 83.6 |