AfriEconQA: leaderboard
Document-grounded question answering: factoid, list, multiple-choice, synthesis and comparison questions over 220 World Bank economic reports on African economies from 2024 and 2025, whose answers are tied to a country, fiscal period and projection state; AfriEconQA (Ajayi, Alaba and Stephen, 2026). The generator answers the 862 held-out test questions from the top five chunks of a fixed hybrid BM25 and BGE-M3 retriever; token F1 in percent (set F1 for list questions), higher is better.
Metric: Token F1 (%; 862 held-out test questions, answered from the top five chunks of a fixed hybrid BM25 + BGE-M3 retriever with reciprocal rank fusion; set F1 for list questions, option letter for multiple choice). Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Pro | 54.5 | #181 |
| 2 | Qwen 3.6 35B A3B | 52.5 | #461 |
| 3 | Gemma 4 12B (IT) | 50.9 | #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 · How It Works · Data refreshed daily, snapshot 2026-11-12.