HELM AFR - Overall (Amharic): leaderboard
Metric: Mean score. Source: crfm.stanford.edu. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 72.18 |
| 2 | Claude 3.7 Sonnet (20250219) | 71.42 |
| 3 | Gemini 2.0 Flash Lite | 66.75 |
| 4 | DeepSeek V3 | 62.83 |
| 5 | Claude 3.5 Haiku (20241022) | 57.35 |
| 6 | Llama 3.1 70B Instruct | 53.06 |
| 7 | GPT-4o (2024-08-06) | 49.79 |
| 8 | Qwen 2.5 72B Instruct | 48.06 |
| 9 | Gemma 2 27B (IT) | 47.53 |
| 10 | Gemma 2 9B (IT) | 41.62 |
| 11 | GPT-4o Mini (2024-07-18) | 40.17 |
| 12 | Mistral 7B Instruct (v0.3) | 35.77 |
| 13 | GPT-3.5 Turbo (0125) | 35.75 |
| 14 | Llama 3.1 8B Instruct | 35.47 |
| 15 | GPT-4 Turbo | 31.65 |
Interactive version: theaggregate.ai/benchmark?slug=helm-afr-overall-amharic · How It Works · Data refreshed daily, snapshot 2026-09-08.