Acropolis

Kimi K2 Thinking

Moonshot AIGrade C. Evidence grade C of A to D: thin.Open weightsReleased 6 Nov 2025
Compare with peers

Benchmark appearances

Best verified run per board, with confidence intervals and Pareto frontier status.

BenchmarkRankScoreCostStatusObserved
FrontierMath-2025-02-28-Private

Epoch AI Benchmarking Hub · Default

#37 / 101
21.4%[16.6% to 26.2%]UnknownStandard5 Dec 2025
Terminal Bench

Epoch AI Benchmarking Hub · Terminus 2

#114 / 202
35.7%[30.2% to 41.3%]UnknownStandard6 Nov 2025
Terminal Bench

Epoch AI Benchmarking Hub · Terminus 2

#115 / 202
35.7%[30.2% to 41.2%]UnknownStandard11 Nov 2025