CursorBench 3.2

We evaluate agents on ambiguous, multi-file tasks from real Cursor sessions. Higher scores are better.

More about CursorBench
A scatter and line chart comparing Fable 5, Opus 4.8, GPT-5.5, Sonnet 5, Grok 4.5, GLM 5.2, Composer 2.5, Gemini 3.5 Flash, and Kimi K2.7 Code scores against average cost per task.75%CursorBench 3.2 score70%65%60%55%50%45%$20$16$12$8$4$0Average cost per taskGrok 4.5 high*Fable 5 highOpus 4.8 highSonnet 5 highComposer 2.5GPT-5.5 mediumGLM 5.2 highKimi K2.7 CodeGemini 3.5 Flash
Model
1Fable 5 Max70.5%$17.32103,52572
2Fable 5 Extra High68.4%$11.7364,97156
3Grok 4.5 High*66.7%$1.5119,52133
4Fable 5 High66.5%$8.7743,74748
5Grok 4.5 Medium*65.4%$1.5418,91434
6Fable 5 Medium65.2%$6.8030,36641
7Grok 4.5 Low*63.5%$1.2215,84131
8Opus 4.8 Max62.3%$5.7771,41144
9Fable 5 Low62.1%$4.4618,18231
10Sonnet 5 Max61.5%$6.4592,88286
11Opus 4.8 Extra High59.4%$4.5051,12140
12Sonnet 5 Extra High58.7%$4.1652,87167
13GPT-5.5 High58.4%$2.0512,18328
14GPT-5.5 Extra High58.4%$2.8517,53432
15Opus 4.8 High58.0%$3.1533,54833
16Sonnet 5 High56.9%$3.1939,48357
17Opus 4.8 Medium56.1%$2.8128,38432
18Composer 2.556.1%$0.4414,28633
19GLM 5.2 Max55.0%$1.7635,94658
20GPT-5.5 Medium53.8%$1.518,52225
21Opus 4.8 Low53.1%$2.0219,62427
22Sonnet 5 Medium52.4%$2.1626,20046
23GLM 5.2 High51.5%$1.1921,82949
24Kimi K2.7 Code49.7%$1.4331,24758
25Gemini 3.5 Flash48.8%$2.2046,70277
26Sonnet 5 Low47.7%$1.3016,26933
27GPT-5.5 Low46.6%$0.985,16820

Grok 4.5 has an advantage on CursorBench: an earlier snapshot of the Cursor codebase was unintentionally included in training. The exact score impact is unclear. That data has been removed for future models. For a rundown of third-party benchmark scores, see the Grok 4.5 launch blog.

Changelog

CursorBench 3.2

  • Introduced instruction following and advanced tool use problems.

CursorBench 3.1

  • Introduced problems focused on codebase understanding, bugfinding, planning, and code review.
  • Improved grading criteria for some edit tasks.

CursorBench 3.0

  • Initial set of tasks focused on edit, refactor, and bugfix problems.

Avg cost / task is computed by applying each model's published per-million-token pricing (input, cache read, cache write, and output) to the tokens it used on each task across the CursorBench 3.2 benchmarks, then averaging with the same task weights as the CursorBench 3.2 score. Results are subject to variance; small differences in scores may not be statistically meaningful.