FrontierScience

A benchmark for research-level scientific reasoning, designed to separate frontier models on difficult science tasks that mix domain knowledge with deep reasoning.

  • Not in index
  • Knowledge
ModelsModels with a result.1
Top resultBest result on this test.N/A
Top-3 spreadPoints from first to third.N/A
YearYear of release.N/A

Results

1 result
#ModelResultThe score from the source.
GPT-5.4 Pro36.7%