⌕Is ___ SOTA?
benchmarks / frontiermath

FrontierMath

accuracy. Status display_only. Contamination risk low. Maintainer Epoch AI. Source.

Display only.This benchmark cannot decide a crown under methodology v1.0.
ModelScoreDateKind
Claude Opus 4.80.47241379310344832026-05-28observation
Qwen3 6 Max Preview0.231034482758620682026-04-20observation
Gemini 3.5 Flash0.38965517241379312026-05-19observation
Qwen3 5 Plus0.21034482758620692026-02-16observation
Qwen3 6 Plus0.26206896551724142026-03-31observation
Qwen3 6 Flash0.103448275862068962026-04-27observation
Qwen3 5 Flash0.062068965517241382026-02-25observation
GLM 5.10.334482758620689662026-04-07observation
Kimi K2 60.38965517241379312026-04-20observation
GPT-5.5 Pro Pre Release0.512026-04-23observation
GPT-5.5 Pro Pre Release0.5242026-04-23observation
GPT-5.5 Pre Release0.5172026-04-23observation
Claude Opus 4.70.43793103448275862026-04-16observation
GPT-5.4 Mini0.28282026-03-17observation
GPT-5.4 Nano0.25862026-03-17observation
Muse Spark0.392026-04-08observation
GPT-5.40.4762026-03-05observation
GPT-5.4 Pro0.52026-03-05observation
Claude Sonnet 4.60.3242026-02-17observation
GLM 50.164335664335664322026-02-11observation
Gemini 3.1 Pro Preview0.3692026-02-19observation
Claude Opus 4.60.4072026-02-05observation
Claude Opus 4.60.396551724137931052026-02-05observation
Fireworks/kimi K2p50.2792026-01-27observation
Zai Org/glm 4.70.0243902439024390252025-12-22observation
Fireworks/deepseek V3p20.2212025-12-01observation
Gemini 2.5 Flash0.048442906574394462025-06-17observation
Gemini 3 Flash Preview0.3564013840830452025-12-17observation
GPT-5.20.4072025-12-11observation
GPT-5.20.403448275862068942025-12-11observation