benchgap
Coding

CursorBench 3.2 leaderboard

As of 2026-11-21, the highest measured score on CursorBench 3.2 is 73.4% by Claude Fable 5.1. 178 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Gemini 4 Neon83.4%estimated ± 4.1 pp, low confidence
2GPT-6.2 Astra82.5%estimated ± 4.1 pp, low confidence
3Gemini 4 Carbon77.4%estimated ± 2.0 pp, medium confidence
4Claude Fable 5.277.2%estimated ± 2.0 pp, medium confidence
5Gemini 4 Hydrogen74.6%estimated ± 2.0 pp, medium confidence
6Gemini 4 Argon73.9%estimated ± 2.0 pp, medium confidence
7Claude Fable 5.173.4%measured
8Muse Wonder72.6%estimated ± 2.0 pp, high confidence
9Claude Mythos 571.4%estimated ± 3.6 pp, high confidence
10DeepSeek V4.1 Flash 101771.1%estimated ± 2.0 pp, high confidence
11Grok 4.670.8%measured
12Claude Fable 570.5%measured
13Claude Opus 5.570.5%estimated ± 1.6 pp, low confidence
14Grok 4.770.2%estimated ± 2.0 pp, high confidence
15Claude Opus 570.0%measured
16GPT-6.1 Terra69.9%estimated ± 3.1 pp, medium confidence
17GPT-6 Astra69.5%estimated ± 1.6 pp, medium confidence
18Gemini 3.8 Flash69.2%measured
19Sakana Fugu-Ultra69.1%estimated ± 3.6 pp, high confidence
20GPT-6.2 Sol68.5%estimated ± 3.1 pp, medium confidence
21GPT-6.1 Sol68.2%estimated ± 3.1 pp, medium confidence
22Ember-168.2%estimated ± 6.4 pp, low confidence
23Pareto 26.968.1%estimated ± 6.4 pp, low confidence
24MAI Code 268.0%estimated ± 6.4 pp, low confidence
25Particle-167.9%estimated ± 6.4 pp, low confidence
26GLM-5.367.9%estimated ± 2.0 pp, high confidence
27DeepSeek V4 Flash 073167.8%estimated ± 3.3 pp, medium confidence
28Muse Wonder Full67.6%estimated ± 3.1 pp, medium confidence
29Pareto 26.10 Preview67.6%estimated ± 6.4 pp, low confidence
30Muse Spark 1.367.2%estimated ± 2.0 pp, high confidence
31GPT-5.6 Sol67.2%measured
32Muse Spark 1.267.1%estimated ± 3.3 pp, medium confidence
33Grok 4.566.7%measured
34SWE-266.7%estimated ± 1.6 pp, medium confidence
35Ling 3.1 Flash66.5%estimated ± 3.6 pp, high confidence
36Ultramarine 1 Flash65.7%estimated ± 6.4 pp, low confidence
37GPT-6 Sol65.6%estimated ± 3.1 pp, medium confidence
38Step 5 Preview65.3%estimated ± 3.6 pp, medium confidence
39MAI Code 2 Flash65.1%estimated ± 6.4 pp, low confidence
40Qwen3.8-27B64.9%estimated ± 3.3 pp, medium confidence
41GPT-5.6 Terra64.9%measured
42Beam 1.0 501B-A23B64.5%estimated ± 3.6 pp, high confidence
43Beam 1.0 501B-A23B FP864.3%estimated ± 3.6 pp, high confidence
44Qwen3.8 Max64.2%estimated ± 3.3 pp, medium confidence
45Hy4 preview64.2%estimated ± 3.6 pp, high confidence
46Beam 1.0 501B-A23B NVFP464.1%estimated ± 3.6 pp, high confidence
47GLM-5.3-Flash63.7%estimated ± 2.0 pp, high confidence
48Ornith-1.5-397B63.7%estimated ± 3.6 pp, high confidence
49Claude Haiku 5.563.4%estimated ± 1.6 pp, medium confidence
50Claude Sonnet 5.563.2%estimated ± 1.6 pp, medium confidence
51SWE-2 Flash62.5%estimated ± 1.6 pp, medium confidence
52Claude Opus 4.862.3%measured
53Inkling 2 Small62.2%estimated ± 4.1 pp, medium confidence
54Qwen3.8-Omni-Flash62.0%estimated ± 3.6 pp, high confidence
55Claude Haiku 4.561.5%estimated ± 3.3 pp, medium confidence
56Claude Sonnet 561.5%measured
57Qwen3.8-Flash-Next61.1%estimated ± 3.6 pp, high confidence
58GPT-5.6 Luna61.1%measured
59Kimi K3.1 Pro61.1%estimated ± 4.1 pp, medium confidence
60Mistral Large 461.0%estimated ± 4.1 pp, medium confidence
61Kimi K360.8%measured
62Ornith-1.0-397B60.8%estimated ± 3.6 pp, high confidence
63Gemini 3.7 Flash60.7%estimated ± 1.6 pp, medium confidence
64Inkling 2 Mini60.2%estimated ± 4.1 pp, low confidence
65Muse Spark 1.160.0%estimated ± 3.6 pp, high confidence
66Muse Spark Open60.0%estimated ± 4.1 pp, low confidence
67Ripple-65B Preview59.9%estimated ± 3.6 pp, high confidence
68Qwen3.8 Max Preview59.7%estimated ± 4.4 pp, high confidence
69Sage-A1.559.6%estimated ± 3.6 pp, high confidence
70SWE-1.759.4%estimated ± 1.6 pp, low confidence
71dots3-note Preview59.4%estimated ± 3.6 pp, high confidence
72GPT-5.1-Codex59.1%estimated ± 7.2 pp, low confidence
73MiniMax M3.1 Flash59.1%estimated ± 5.3 pp, medium confidence
74o359.1%estimated ± 7.2 pp, low confidence
75GPT-5.1-Codex-Max59.1%estimated ± 7.2 pp, low confidence
76Grok 459.1%estimated ± 7.2 pp, low confidence
77o4-mini59.1%estimated ± 7.2 pp, low confidence
78GLM-4.659.1%estimated ± 7.2 pp, low confidence
79Grok 4.1 Fast (Reasoning)59.1%estimated ± 7.2 pp, low confidence
80Grok 4 Fast (Reasoning)59.1%estimated ± 7.2 pp, low confidence
81Qwen3 Max59.1%estimated ± 7.2 pp, low confidence
82o3-mini59.1%estimated ± 7.2 pp, low confidence
83Claude 3.5 Sonnet59.1%estimated ± 7.2 pp, low confidence
84Claude 3.7 Sonnet59.1%estimated ± 7.2 pp, low confidence
85Claude 3.7 Sonnet Thinking59.1%estimated ± 7.2 pp, low confidence
86GLM-4.559.1%estimated ± 7.2 pp, low confidence
87GPT-4.159.1%estimated ± 7.2 pp, low confidence
88GPT-4.1 mini59.1%estimated ± 7.2 pp, low confidence
89Grok Code Fast 159.1%estimated ± 7.2 pp, low confidence
90o159.1%estimated ± 7.2 pp, low confidence
91Qwen3.7 Max58.9%estimated ± 3.6 pp, high confidence
92GPT-5.558.4%measured
93Atria Dawn Preview57.6%estimated ± 3.6 pp, high confidence
94Ornith-1.5-35B-A3B57.6%estimated ± 3.6 pp, high confidence
95Laguna S 2.157.3%estimated ± 3.6 pp, high confidence
96Claude Opus 4.6 (Adaptive)57.2%estimated ± 5.4 pp, low confidence
97MiniMax M356.8%estimated ± 3.6 pp, high confidence
98Sakana Fugu56.8%estimated ± 3.6 pp, high confidence
99Claude Opus 4.5 Thinking56.2%estimated ± 5.4 pp, low confidence
100Gemini 3 Pro56.2%estimated ± 5.4 pp, low confidence
101Kimi K2.656.2%estimated ± 3.6 pp, high confidence
102Composer 2.556.1%measured
103GLM-5.155.9%estimated ± 3.6 pp, high confidence
104Gemini 3.1 Pro55.8%estimated ± 4.4 pp, high confidence
105Claude Opus 4.755.6%estimated ± 1.6 pp, low confidence
106Gemini 3 Flash55.5%estimated ± 5.4 pp, low confidence
107GLM-5.255.0%measured
108GPT-5.454.9%estimated ± 3.6 pp, high confidence
109Qwen3.7 Plus54.7%estimated ± 3.6 pp, high confidence
110Qwen 3.6 Max (preview)54.3%estimated ± 3.6 pp, high confidence
111MiMo-V2.5-Pro54.1%estimated ± 3.6 pp, high confidence
112GPT-5.2-Codex54.1%estimated ± 5.4 pp, low confidence
113Claude Opus 4.553.9%estimated ± 3.6 pp, high confidence
114Gemini 3.6 Flash53.5%measured
115GPT-5.3 Codex53.5%estimated ± 3.6 pp, high confidence
116Mercury 353.4%estimated ± 6.4 pp, low confidence
117Ling 3.0 Flash53.1%estimated ± 3.6 pp, high confidence
118Qwen3.6 Plus53.1%estimated ± 3.6 pp, high confidence
119Claude Sonnet 4.5 Thinking52.7%estimated ± 5.4 pp, low confidence
120Step 3.7 Flash52.6%estimated ± 3.6 pp, high confidence
121MiniMax M2.752.5%estimated ± 3.6 pp, high confidence
122MiMo-V2.552.3%estimated ± 3.6 pp, high confidence
123GPT-552.2%estimated ± 5.4 pp, low confidence
124Inkling-Small52.0%estimated ± 3.6 pp, high confidence
125GPT-5.251.4%estimated ± 3.6 pp, high confidence
126DeepSeek V4 Pro51.1%estimated ± 3.6 pp, high confidence
127Claude Sonnet 4.651.0%estimated ± 5.0 pp, low confidence
128GLM-550.6%estimated ± 3.6 pp, high confidence
129Solar Pro 450.5%estimated ± 5.3 pp, medium confidence
130Kimi K2.7 Code49.7%measured
131Qwen3.5 Flash49.6%estimated ± 5.4 pp, low confidence
132Inkling49.1%estimated ± 3.6 pp, medium confidence
133Spark-X2.6-4B49.1%estimated ± 3.6 pp, medium confidence
134Gemini 3.5 Flash-Lite48.9%estimated ± 3.6 pp, medium confidence
135Gemini 3.5 Flash48.8%measured
136Gemini 3.1 Flash-Lite48.6%estimated ± 5.4 pp, low confidence
137GPT-6 Luna48.0%estimated ± 3.1 pp, low confidence
138Gemini 2.5 Pro47.8%estimated ± 5.3 pp, low confidence
139Qwen3.6-27B47.6%estimated ± 3.6 pp, medium confidence
140Claude Opus 4.647.4%estimated ± 3.6 pp, medium confidence
141Quasar 438B46.5%estimated ± 4.4 pp, high confidence
142MAI-Thinking-146.2%estimated ± 3.6 pp, medium confidence
143Apodex 1.146.0%estimated ± 4.4 pp, high confidence
144Apodex 1.1 Mini46.0%estimated ± 4.4 pp, high confidence
145Muse Spark45.4%estimated ± 3.6 pp, medium confidence
146Ling 3.0 Flash VL44.4%estimated ± 5.3 pp, low confidence
147Grok 4.2044.1%estimated ± 3.6 pp, medium confidence
148Nemotron 3.5 Super VL 120B A12B44.1%estimated ± 3.6 pp, medium confidence
149Hy343.9%estimated ± 4.4 pp, medium confidence
150Hy3 Preview43.9%estimated ± 4.4 pp, medium confidence
151Nemotron 3.5 Super VL 120B A12B NVFP443.1%estimated ± 3.6 pp, medium confidence
152Muse Glimmer 30B42.8%estimated ± 3.6 pp, medium confidence
153Qwen3.5 397B42.2%estimated ± 3.6 pp, medium confidence
154Kimi K2.541.8%estimated ± 3.6 pp, medium confidence
155Ornith-1.0-35B41.1%estimated ± 3.6 pp, medium confidence
156GPT-5.4 mini41.0%estimated ± 4.4 pp, medium confidence
157GPT-5.4 nano41.0%estimated ± 4.4 pp, medium confidence
158K-EXAONE 2.040.9%estimated ± 5.3 pp, low confidence
159A.X K239.3%estimated ± 5.3 pp, low confidence
160Qwen3.6-35B-A3B39.1%estimated ± 3.6 pp, medium confidence
161Trinity-Large-Preview38.7%estimated ± 5.3 pp, low confidence
162Trinity-Large-Thinking38.7%estimated ± 5.3 pp, low confidence
163Laguna M.138.4%estimated ± 3.6 pp, medium confidence
164Gemma 4 26B A4B37.7%estimated ± 5.3 pp, low confidence
165Nemotron 3.5 Bolt 74B-A9B37.2%estimated ± 5.3 pp, low confidence
166DeepSeek V3 032436.1%estimated ± 5.3 pp, low confidence
167GPT-OSS 20B36.0%estimated ± 5.3 pp, low confidence
168Mistral Small 435.8%estimated ± 5.3 pp, low confidence
169North Mini Code35.8%estimated ± 5.3 pp, low confidence
170Command A+35.3%estimated ± 5.3 pp, low confidence
171Mercury 2.535.3%estimated ± 5.3 pp, low confidence
172MiMo-V2-Flash34.7%estimated ± 4.4 pp, medium confidence
173Laguna XS 2.134.7%estimated ± 3.6 pp, medium confidence
174Ornith-1.5-9B34.5%estimated ± 3.6 pp, medium confidence
175GPT-5.134.3%estimated ± 4.4 pp, medium confidence
176Nemotron 3 Ultra34.2%estimated ± 4.4 pp, medium confidence
177Mistral Large 332.3%estimated ± 5.3 pp, low confidence
178Mistral Medium 3.5 128B32.0%estimated ± 4.4 pp, medium confidence
179Qwen4 Max31.9%estimated ± 3.6 pp, low confidence
180Laguna XS.231.6%estimated ± 3.6 pp, medium confidence
181DeepSeek V331.0%estimated ± 5.3 pp, low confidence
182Qwen3.5-122B-A10B30.9%estimated ± 4.4 pp, medium confidence
183Hy430.6%estimated ± 3.6 pp, low confidence
184GLM-4.730.5%estimated ± 4.4 pp, medium confidence
185ERNIE 5.5 Flash Preview30.2%estimated ± 3.6 pp, low confidence
186Gemma 4 31B28.9%estimated ± 4.4 pp, medium confidence
187GPT-OSS 120B28.1%estimated ± 5.3 pp, low confidence
188Grok 4.327.9%estimated ± 4.4 pp, medium confidence
189MiMo-V2.6-Pro27.5%estimated ± 3.6 pp, low confidence
190MiMo-V2.6-Flash27.1%estimated ± 3.6 pp, low confidence
191LongCat-2.526.6%estimated ± 3.6 pp, low confidence
192Ornith-1.0-9B23.7%estimated ± 3.6 pp, medium confidence
193LongCat-Flash-Lite-Sparse18.7%estimated ± 3.6 pp, medium confidence
194Hy4 flash17.6%estimated ± 3.6 pp, low confidence
195Granite 4.2 30B6.9%estimated ± 3.6 pp, medium confidence
196LLaDA2.2-flash3.9%estimated ± 3.6 pp, medium confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General