MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark

MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark is an AI benchmark in the Benchmark Radar catalog.

Released: 2024-06-04
Openness: unknown
Importer: Benchmark Saturation

Source provenance

Original evidence

MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark results and reported scores

Reported benchmark scores · Benchmark Saturation

Reported scores: 5

Model	Organization	Reported value	Reported	Evidence
Claude Opus 4	—	86.2	—	evidence
Claude Opus 4.1	—	87.8	—	evidence
Claude Sonnet 4.5	—	87.3	—	evidence
GPT 5	—	87.0	—	evidence
GPT o1	—	89.3	—	evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.

Which sources cite MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark?