CMRC (Chinese Machine Reading Comprehension) — Chinese reading-comprehension benchmark, EM metric
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | DeepSeek-V2 | DeepSeek | 77.4 | 1 | 2026-05-01 |
| 2 | DeepSeek-V3 | DeepSeek | 76.3 | 1 | 2026-05-01 |
| 3 | Llama 3.1 Instruct 405B | Meta | 76 | 1 | 2026-05-01 |
| 4 | Qwen2.5 Instruct 72B | Alibaba | 75.8 | 1 | 2026-05-01 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.