IyawoBench: leaderboard
IyawoBench (Iyawo Health, 2026) tests LLM clinical triage of febrile illness in Nigerian primary care on 200 synthetic vignettes derived from real primary health centre encounters, each labelled treat here, refer today or refer now under WHO and Nigerian guidelines, at temperature 0. The score is three-way triage accuracy in percent; higher is better.
Metric: Three-way triage accuracy (%) over the 200 IyawoBench v1.0 vignettes (100 refer now, 60 refer today, 40 treat here), a triage decision with a one-sentence rationale at temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.3 70B Instruct | 68 | #1357 |
| 2 | Claude Sonnet 4.6 | 63 | #161 |
| 3 | Llama 3.1 8B Instruct | 46 | #4246 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=iyawobench · How It Works · Data refreshed daily, snapshot 2026-11-13.