aiaimodel.cards
← All benchmarks

MMMU (validation)

College-level questions that require reading images, charts and diagrams.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Claude Opus 4.5 · Anthropicbest value
    80.7%

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Claude Opus 4.5 Table 2.3.A, p.19

    Average of 5 trials with a 64k thinking budget, 200k context, default (high) effort and default sampling. SWE-bench, τ²-bench and MCP Atlas are without extended thinking; Terminal-bench uses a 128k budget (57.8% at 64k). GPT-5.1 MMMU is from the MMMU leaderboard.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.