Skip to content
Bencher

TestsBenchmarks · Coding

CursorBench

Share of ambiguous, multi-file tasks drawn from real Cursor sessions that an agent completes inside Cursor, per model and effort level.

% solvedHigher is betterCounts toward:Weights: Coding ×0.9The test's websiteOfficial page ↗

The best 15 · 15 models tested

Top 15 · best setting per model · 15 models, 63 results (63 independent)

Measured by independent testersIndependent Reported by the makerVendor-reported
CursorBench leaderboard
Thinking levelSetting
1Claude Opus 5.5Anthropic57.8%MaxOpus 5.5 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
2Claude Sonnet 5.5Anthropic55.5%MaxSonnet 5.5 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
3Claude Fable 5.1Anthropic51.8%MaxFable 5.1 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
4Claude Opus 5Anthropic46.6%MaxOpus 5 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
5Grok 4.7SpaceXAI (formerly xAI)46.3%Extra highGrok 4.7 Extra HighIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
6GLM-5.3Z.ai (Zhipu)42.6%MaxGLM 5.3 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
7GPT-5.6 SolOpenAI41.7%MaxGPT-5.6 Sol MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
8Muse Spark 1.3Meta (Meta Superintelligence Labs, Muse)41.6%MaxMuse Spark 1.3 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
9Grok 4.6SpaceXAI (formerly xAI)41.4%Extra highGrok 4.6 Extra HighIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
10GPT-5.6 TerraOpenAI41.3%MaxGPT-5.6 Terra MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
11Gemini 3.8 FlashGoogle (Gemini / DeepMind)39.6%HighGemini 3.8 Flash HighIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
12GLM-5.3-FlashZ.ai (Zhipu)36.8%MaxGLM 5.3 Flash MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
13GPT-5.6 LunaOpenAI35.9%MaxGPT-5.6 Luna MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
14Claude Sonnet 5Anthropic34.1%MaxSonnet 5 MaxIndependent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026
15Composer 2.5Cursor27.7%DefaultComposer 2.5Independent testIndependentCursor (CursorBench) ↗Cursor agent1 Oct 2026