Wikibench

IFEval

general

Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints

Leaderboard

Showing 20 of 64 results

  1. Qwen3.5-27B
    95.0%
    i
  2. Qwen3.6 Plus
    94.3%
    i
  3. Qwen3.7 Max
    94.3%
    i
  4. o3-mini
    93.9%
    i
  5. Qwen3.5-122B-A10B
    93.4%
    i
  6. Claude 3.7 Sonnet
    93.2%
    i
  7. Qwen3.5-397B-A17B
    92.6%
    i
  8. Nova Pro
    92.1%
    i
  9. Llama 3.3 70B Instruct
    92.1%
    i
  10. Qwen3.5-35B-A3B
    91.9%
    i
  11. Qwen3.5-9B
    91.5%
    i
  12. Gemma 3 27B
    90.4%
    i
  13. Nemotron Nano 9B v2
    90.3%
    i
  14. Gemma 3 4B
    90.2%
    i
  15. Kimi K2 Instruct
    89.8%
    i
  16. Kimi K2-Instruct-0905
    89.8%
    i
  17. Qwen3.5-4B
    89.8%
    i
  18. Nova Lite
    89.7%
    i
  19. LongCat-Flash-Chat
    89.6%
    i
  20. Llama 3.1 Nemotron Ultra 253B v1
    89.5%
    i