# Chinese Massive Multitask Language Understanding (CMMLU)

> A Chinese multitask academic benchmark reported in DeepSeek-V4 base-model evaluations.

Canonical page: https://benchlm.ai/benchmarks/cmmlu

- Category: [Knowledge](/knowledge)
- Last updated: September 15, 2026

## About CMMLU

- Year: 2026
- Tasks: Chinese academic QA
- Format: Exact match
- Difficulty: Broad Chinese knowledge
- Paper: [DeepSeek-V4 Technical Report](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/blob/main/DeepSeek_V4.pdf)

BenchLM stores CMMLU as a display-only provider-table row when exact values are published in DeepSeek-V4 evaluations.

CMMLU is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (1 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [LongCat-Flash-Lite-Sparse](/models/longcat-flash-lite-sparse) | Meituan | 84.3% |

## FAQ

### What does CMMLU measure?

A Chinese multitask academic benchmark reported in DeepSeek-V4 base-model evaluations.

### Which model scores highest on CMMLU?

LongCat-Flash-Lite-Sparse by Meituan currently leads with a score of 84.3% on CMMLU.

### How many models are evaluated on CMMLU?

1 AI models have been evaluated on CMMLU on BenchLM.

### Does CMMLU affect BenchLM's overall score?

Not directly. CMMLU is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.