Skip to main content

CMath

A Chinese mathematical reasoning benchmark reported in DeepSeek-V4 base-model evaluations.

Benchmark score on CMath — April 24, 2026

BenchLM mirrors the published score view for CMath. DeepSeek V4 Flash Base leads the public snapshot at 93.6% , followed by DeepSeek V4 Pro Base (90.9%). BenchLM does not use these results to rank models overall.

2 modelsMathCurrentDisplay onlyUpdated April 24, 2026

About CMath

Year

2026

Tasks

Chinese math problems

Format

Exact match

Difficulty

Math reasoning

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

BenchLM freshness & provenance

Version

CMath 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Benchmark score table (2 models)

1
93.6%
2
90.9%

FAQ

What does CMath measure?

A Chinese mathematical reasoning benchmark reported in DeepSeek-V4 base-model evaluations.

Which model scores highest on CMath?

DeepSeek V4 Flash Base by DeepSeek currently leads with a score of 93.6% on CMath.

How many models are evaluated on CMath?

2 AI models have been evaluated on CMath on BenchLM.

Last updated: April 24, 2026 · BenchLM version CMath 2026

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.