MMMLU

Citation frequency 2/3
What does it measure?
A multilingual test of broad knowledge across many languages, released by OpenAI.
Who evaluates it, and how?
MMLU questions (a multiple-choice knowledge exam covering 57 subjects) translated into 14 languages by professional translators.
Reading the score
Scores run from 0 to 100%. Higher is better.

This ranking is built only from scores the model makers published themselves. Test settings such as tool use and reasoning effort differ by maker, so check each score's setting before comparing models directly.

Top score92.6%
Models tested18
Scoring basisHigher is better
Last updated2026.08.05
Model release date
  1. 1
    Google2026.02.19 · Thinking (High)
    92.6%
  2. 2
    Google2025.12.17 · Thinking
    91.8%
  3. 3
    Anthropic2026.04.16 · max (adaptive thinking)
    91.5%
  4. 4
    Anthropic2026.02.04 · max (adaptive thinking)
    91.1%
  5. 5
    Anthropic2025.11.24 · 64k thinking budget, default effort (high)
    90.8%
  6. 6
    Alibaba2026.05.21
    90.3%
  7. 7
    Alibaba2026.04.02
    89.5%
  8. 7
    Google2025.06.17 · Thinking
    89.5%
  9. 9
    Anthropic2026.02.17 · max (adaptive thinking)
    89.3%
  10. 10
    Anthropic2025.09.29 · extended thinking (up to 128K)
    89.1%

The "harness" is the agent program the AI used to carry out the task. The same model can score very differently depending on its harness and reasoning effort.

  • Self-reported by each model maker — see the source link on each scoreFigures that model makers announced themselves, checked against the original post or system card and copied over. Conditions differ by maker, and each score links to its original source.