HELM AFR - MMLU Clinical Knowledge (Shona): leaderboard
Metric: EM. Source: crfm.stanford.edu. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 81.13 |
| 2 | Claude 3.7 Sonnet (20250219) | 77.36 |
| 3 | GPT-4o (2024-08-06) | 76.6 |
| 4 | Gemini 2.0 Flash Lite | 73.58 |
| 5 | DeepSeek V3 | 72.08 |
| 6 | GPT-4 Turbo | 66.42 |
| 7 | Llama 3.1 405B Instruct | 60.75 |
| 8 | Llama 3.1 70B Instruct | 55.47 |
| 9 | Claude 3.5 Haiku (20241022) | 53.21 |
| 10 | Llama 3.3 70B Instruct | 52.45 |
| 11 | Gemma 2 27B (IT) | 52.45 |
| 12 | Qwen 2.5 72B Instruct | 46.79 |
| 13 | GPT-3.5 Turbo (0125) | 44.15 |
| 14 | Gemma 2 9B (IT) | 41.51 |
| 15 | Llama 3.1 8B Instruct | 39.25 |
Interactive version: theaggregate.ai/benchmark?slug=helm-afr-mmlu-clinical-knowledge-shona · How It Works · Data refreshed daily, snapshot 2026-09-08.