Cursorbench

Models scored

31

Best score

0.73

Metric

Score

Cursorbench is an AI evaluation tracked by Epoch AI, with 31 scored model versions on record. The scores here are reported by third parties rather than produced by Epoch AI running the evaluation itself.

Scored models span January 2026 to July 2026. The best score recorded is 0.73, measured as Score. Models come from Anthropic, OpenAI, Z.ai (Zhipu AI), Google, Moonshot.

Full record

Score metric: Score
Models scored: 31
Best score recorded: 0.73
Earliest model scored: Jan 27, 2026
Most recent model scored: Jul 24, 2026
Third-party reported: Yes

Leaderboard

01	claude-fable-5_max	Anthropic	0.73
02	claude-fable-5_xhigh	Anthropic	0.72
03	claude-fable-5_high	Anthropic	0.71
04	claude-opus-5_max	Anthropic	0.7
05	gpt-5.6-sol_max	OpenAI	0.67
06	claude-opus-4-7_max	Anthropic	0.65
07	gpt-5.5_xhigh	OpenAI	0.64
08	claude-opus-4-8_max	Anthropic	0.64
09	gpt-5.5_high	OpenAI	0.63
10	claude-opus-4-8_xhigh	Anthropic	0.62
11	claude-opus-4-7_xhigh	Anthropic	0.62
12	claude-sonnet-5_max	Anthropic	0.61
13	claude-opus-4-7_high	Anthropic	0.59
14	gpt-5.5_medium	OpenAI	0.59
15	claude-sonnet-5_xhigh	Anthropic	0.58
16	claude-opus-4-8_high	Anthropic	0.58
17	claude-opus-4-8_medium	Anthropic	0.57
18	glm-5.2_max	Z.ai (Zhipu AI)	0.55
19	claude-opus-4-8_low	Anthropic	0.54
20	claude-opus-4-7_medium	Anthropic	0.53
21	glm-5.2_high	Z.ai (Zhipu AI)	0.51
22	gemini-3.5-flash_unknown	Google	0.5
23	claude-sonnet-4-6_max	Anthropic	0.49
24	claude-sonnet-4-6_high	Anthropic	0.49
25	gpt-5.5_low	OpenAI	0.49
26	claude-opus-4-7_low	Anthropic	0.48
27	kimi-k2.6	Moonshot	0.48
28	claude-sonnet-4-6_medium	Anthropic	0.46
29	kimi-k2.5	Moonshot	0.32

More benchmarks

Epoch Capabilities Index MMLU Terminal-Bench GPQA Diamond OTIS Mock AIME 2024-2025 GSM8K ARC-AGI ARC-AGI-2

SourceEpoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/benchmarks. Licensed under CC BY 4.0.

← All benchmarks