AfriEconQA (Oracle Evidence): leaderboard
AfriEconQA (Ajayi, Alaba and Stephen, 2026) with the short gold evidence passage given in place of retrieved chunks, testing whether a model can use the right evidence for exact numerical and temporal answers about African economies. Token F1 in percent over the 862 held-out test questions; higher is better.
Metric: Token F1 (%; 862 held-out test questions, answered from the short gold evidence passage). Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.6 35B A3B | 48.9 | #461 |
| 2 | DeepSeek V4 Pro | 47.3 | #181 |
| 3 | Gemma 4 12B (IT) | 40.3 | #688 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=afrieconqa-oracle-evidence · How It Works · Data refreshed daily, snapshot 2026-11-12.