Wikibench

IMO-AnswerBench

math

Multilingual: no

IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.

Methodology

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

Leaderboard

Showing 14 of 14 results

  1. Qwen3.7 Max self-reported llm-stats
    90.0%
  2. DeepSeek-V4-Pro-Max self-reported llm-stats
    89.8%
  3. DeepSeek-V4-Flash-Max self-reported llm-stats
    88.4%
  4. Kimi K2.6 self-reported llm-stats
    86.0%
  5. Step-3.5-Flash self-reported llm-stats
    85.4%
  6. GLM-5.1 self-reported llm-stats
    83.8%
  7. Qwen3.6 Plus self-reported llm-stats
    83.8%
  8. GLM-4.7 self-reported llm-stats
    82.0%
  9. Kimi K2.5 self-reported llm-stats
    81.8%
  10. Qwen3.5-397B-A17B self-reported llm-stats
    80.9%
  11. Qwen3.6-27B self-reported llm-stats
    80.8%
  12. Kimi K2-Thinking-0905 self-reported llm-stats
    78.6%
  13. LongCat-Flash-Thinking-2601 self-reported llm-stats
    78.6%
  14. DeepSeek-V3.2 self-reported llm-stats
    78.3%