Wikibench

MMLU

math

Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains

Leaderboard

Showing 20 of 105 results

  1. GPT-5
    92.5%
    i
  2. o1
    91.8%
    i
  3. GPT-4.5
    90.8%
    i
  4. o1-preview
    90.8%
    i
  5. Qwen3 VL 235B A22B Thinking
    90.6%
    i
  6. Sarvam-105B
    90.6%
    i
  7. Claude 3.5 Sonnet
    90.4%
    i
  8. Claude 3.5 Sonnet
    90.4%
    i
  9. Kimi K2 0905
    90.2%
    i
  10. GPT-4.1
    90.2%
    i
  11. GPT OSS 120B
    90.0%
    i
  12. LongCat-Flash-Chat
    89.7%
    i
  13. Kimi K2 Instruct
    89.5%
    i
  14. Kimi K2-Instruct-0905
    89.5%
    i
  15. MiMo-V2.5-Pro
    89.4%
    i
  16. Qwen3 VL 235B A22B Instruct
    88.8%
    i
  17. GPT-4o
    88.7%
    i
  18. Claude 3.5 Sonnet
    88.7%
    i
  19. GPT-4o
    88.7%
    i
  20. Qwen3 VL 32B Thinking
    88.7%
    i