AfriScience-MT (10-Shot): leaderboard
Sentence-level translation of scientific summaries between English and six African languages on the AfriScience-MT test split (Abdulmumin et al., 2026), with the paper's shared prompt template and ten random in-context examples at temperature 0. SSA-COMET on a 0-100 scale averaged over the 12 directions; higher is better.
Metric: SSA-COMET (0-100; sentence-level, mean over the 12 English<->African directions of the test split; shared prompt template 5 with 10 random in-context examples, temperature 0). Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 68.3 | #121 |
| 2 | Gemini 3.1 Flash Lite | 68 | #363 |
| 3 | Gemini 1.5 Flash | 65.9 | #1671 |
| 4 | GPT-4o | 65.4 | #741 |
| 5 | Tiny Aya Global | 55.1 | #6489 |
| 6 | Gemma 2 9B (IT) | 51.4 | #2275 |
| 7 | Llama 3.1 8B Instruct | 46.9 | #4277 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=afriscience-mt-10-shot · How It Works · Data refreshed daily, snapshot 2026-11-12.