HELM AFR - MMLU Clinical Knowledge (Afrikaans): leaderboard
Metric: EM. Source: crfm.stanford.edu. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 88.68 |
| 2 | GPT-4o (2024-08-06) | 88.3 |
| 3 | DeepSeek V3 | 86.79 |
| 4 | Claude 3.7 Sonnet (20250219) | 86.79 |
| 5 | Qwen 2.5 72B Instruct | 83.4 |
| 6 | Gemini 2.0 Flash Lite | 80.75 |
| 7 | Llama 3.1 70B Instruct | 78.11 |
| 8 | GPT-4 Turbo | 70.57 |
| 9 | Gemma 2 27B (IT) | 70.19 |
| 10 | Llama 3.1 405B Instruct | 69.06 |
| 11 | Claude 3.5 Haiku (20241022) | 68.68 |
| 12 | GPT-3.5 Turbo (0125) | 66.04 |
| 13 | GPT-4o Mini (2024-07-18) | 59.62 |
| 14 | Gemma 2 9B (IT) | 54.34 |
| 15 | Llama 3.3 70B Instruct | 49.81 |
Interactive version: theaggregate.ai/benchmark?slug=helm-afr-mmlu-clinical-knowledge-afrikaans · How It Works · Data refreshed daily, snapshot 2026-09-08.