MultiNRC
General QA · 2025-07-23
Score on native multilingual reasoning tasks in French, Spanish, and Chinese.
Top models (higher is better)
| Model | Score |
|---|---|
| Muse Spark 1.1 | 65.6 |
| GPT-5 Pro | 65.2 |
| Gemini 3.1 Pro Preview | 64.7 |
| GPT-5.4 Pro | 62.3 |
| Muse Spark | 59.0 |
| Gemini 3 Pro Preview | 59.0 |
| GPT-5.4 | 58.3 |
| Opus 4.6 | 57.1 |
| GPT-5 | 52.1 |
| GPT-5.1 | 49.0 |
| o3-pro | 49.0 |
| Opus 4.5 | 48.6 |
| o3 | 45.5 |
| Gemini 2.5 Pro Preview 06-05 | 45.1 |
| GPT-5.2 | 42.2 |
| Opus 4.1 | 38.4 |
| Sonnet 4.5 | 35.8 |
| Kimi K2.5 | 35.2 |
| Opus 4 | 33.9 |
| Claude 3.7 Sonnet | 27.8 |
| DeepSeek-R1-0528 | 27.6 |
| Qwen3 235B A22B Thinking 2507 | 27.1 |
| Gemini 3.1 Flash-Lite Preview | 25.0 |
| DeepSeek-R1 | 24.3 |
| GPT-5 Mini | 23.9 |
| DeepSeek-V3.1 | 23.6 |
| O4 Mini | 22.2 |
| GPT-4.1 | 21.2 |
| Kimi K2 Instruct | 18.5 |
| Sonnet 4 | 18.4 |
| GPT-5.1 Instant | 18.3 |
| Qwen3-235B-A22B | 17.6 |
| GLM-4.5 | 17.4 |
| gpt-oss-120b | 15.2 |
| GPT-4o | 12.4 |
| GLM 4.5 Air | 10.4 |
| gpt-oss-20b | 10.4 |
| Llama 4 Maverick Instruct | 8.4 |