Wikibench

Internal API instruction following (hard)

general

Internal API instruction following (hard) benchmark - specific documentation not found in official sources

Leaderboard

Showing 7 of 7 results

  1. GPT-5
    64.0%
    i
  2. GPT-4.5
    54.0%
    i
  3. o3-mini
    50.0%
    i
  4. GPT-4.1
    49.1%
    i
  5. GPT-4.1 mini
    45.1%
    i
  6. GPT-4.1 nano
    31.6%
    i
  7. GPT-4o
    29.2%
    i