Wikibench

C-Eval

reasoning official site →

Multilingual: no

C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: general, reasoning. Language: en. Verified by llm-stats: no.

Leaderboard

Showing 12 of 12 results

  1. Qwen3.6 Plus self-reported llm-stats
    93.3%
  2. Qwen3.5-397B-A17B self-reported llm-stats
    93.0%
  3. Kimi K2 Base self-reported llm-stats
    92.5%
  4. Qwen3.5-122B-A10B self-reported llm-stats
    91.9%
  5. Qwen3.6-27B self-reported llm-stats
    91.4%
  6. Qwen3.5-27B self-reported llm-stats
    90.5%
  7. Qwen3.5-35B-A3B self-reported llm-stats
    90.2%
  8. Kimi-k1.5 self-reported llm-stats
    88.3%
  9. DeepSeek-V3 self-reported llm-stats
    86.5%
  10. Qwen2 72B Instruct self-reported llm-stats
    83.8%
  11. Qwen2 7B Instruct self-reported llm-stats
    77.2%
  12. ERNIE 4.5 self-reported llm-stats
    40.7%