CSimpleQA: leaderboard
3,000 short-answer Chinese factual questions across 6 topics and 99 subtopics from Alibaba's Taobao & Tmall Group (2024); an LLM grader marks each correct, incorrect or not attempted; F-score.
Metric: Score (%). Source: arxiv.org. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 41 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O1 Preview | 67.9 | #193 |
| 2 | GLM-4 Plus | 60.9 | #388 |
| 3 | GPT-4o | 59.7 | #343 |
| 4 | Qwen Max | 57.4 | #382 |
| 5 | Gemini 1.5 Pro | 56.7 | #445 |
| 6 | DeepSeek V2.5 | 55.7 | #489 |
| 7 | Claude 3.5 Sonnet | 53.5 | #350 |
| 8 | Yi Large | 51.5 | #565 |
| 9 | Qwen 2.5 72B | 50.2 | #499 |
| 10 | Moonshot v1 8K | 50.1 | #617 |
| 11 | GPT-4 Turbo | 49.1 | #404 |
| 12 | GPT-4 | 47.4 | #438 |
| 13 | O1 Mini | 44.1 | #323 |
| 14 | Qwen 2.5 32B | 41.1 | #546 |
| 15 | Llama 3.1 70B | 40.2 | #592 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=csimpleqa · How It Works · Data refreshed daily, snapshot 2026-10-20.