Skip to content

Apex

A high-difficulty mathematical reasoning benchmark reported in DeepSeek-V4 model evaluations.

8 models

Higher is better.

  1. 1Hy4 previewTencent74.2
  2. 2DeepSeek V4.1 Flash 1017DeepSeek65.6
  3. 3A.X K2SK Telecom45.8
  4. 4Qwen3.7 MaxAlibaba44.5
  5. 5DeepSeek V4 Pro 0813DeepSeek38.3
  6. 6DeepSeek V4 Flash 0731DeepSeek33.0
  7. 7ZAYA1-8BZyphra32.2
  8. 8Qwen3.7 PlusAlibaba22.7
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Pass@1 math benchmark
Tasks
Advanced mathematical reasoning
Authors
DeepSeek-AI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.