Inference
EN

cursorBench40

ModelsModels with a result.21
Top resultBest result on this benchmark.60.4%Gemini 4 Carbon
Top-3 spreadPoints from first to third.2.6 pts
YearYear of release.—

Result and price

21 of 21 results

Results

21 results
#ModelResultThe score from the source.
01Gemini 4 Carbon60.4%
02Claude Fable 5.260.0%
03Claude Opus 5.557.8%
04Claude Sonnet 5.555.5%
05Gemini 4 Argon52.9%
06Claude Fable 5.151.8%
07Muse Wonder50.4%
08Claude Haiku 5.548.4%
09DeepSeek V4.1 Flash 101747.7%
10Claude Opus 546.6%
11Grok 4.746.3%
12GLM-5.342.6%
13GPT-5.6 Sol41.7%
14Muse Spark 1.341.6%
15Grok 4.641.4%
16GPT-5.6 Terra41.3%
17Gemini 3.8 Flash39.6%
18GLM-5.3-Flash36.8%
19GPT-5.6 Luna35.9%
20Claude Sonnet 534.1%
21Composer 2.527.7%