benchgap
Coding

CursorBench 4.0 leaderboard

As of 2026-11-21, the highest measured score on CursorBench 4.0 is 60.4% by Gemini 4 Carbon. 155 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Gemini 4 Neon62.4%estimated ± 2.6 pp, medium confidence
2Gemini 4 Carbon60.4%measured
3Claude Fable 5.260.0%measured
4Claude Opus 5.557.8%measured
5GPT-6.2 Astra57.6%estimated ± 3.3 pp, high confidence
6GPT-6.2 Sol57.4%estimated ± 3.3 pp, high confidence
7Claude Sonnet 5.555.5%measured
8GPT-6.1 Sol54.7%estimated ± 3.3 pp, high confidence
9Gemini 4 Hydrogen54.3%measured
10Step 5 Preview53.9%estimated ± 4.7 pp, high confidence
11Gemini 4 Argon52.9%measured
12Muse Wonder Full51.9%estimated ± 4.7 pp, high confidence
13Claude Fable 5.151.8%measured
14GPT-6 Sol51.6%estimated ± 3.3 pp, high confidence
15Claude Mythos 550.9%estimated ± 3.3 pp, high confidence
16Muse Wonder50.4%measured
17Hy449.4%estimated ± 3.3 pp, high confidence
18MiniMax M3.1 Flash48.6%estimated ± 3.3 pp, medium confidence
19Claude Haiku 5.548.4%measured
20Qwen4 Max48.0%estimated ± 4.7 pp, high confidence
21DeepSeek V4.1 Flash 101747.7%measured
22ERNIE 5.5 Flash Preview47.4%estimated ± 4.7 pp, high confidence
23SWE-247.2%estimated ± 3.3 pp, high confidence
24SWE-2 Flash47.2%estimated ± 3.3 pp, high confidence
25GPT-6 Luna47.0%estimated ± 3.3 pp, high confidence
26Kimi K3.1 Pro46.7%estimated ± 3.3 pp, medium confidence
27Mistral Large 446.7%estimated ± 3.3 pp, medium confidence
28Claude Opus 546.6%measured
29SWE-1.746.5%estimated ± 3.3 pp, medium confidence
30Inkling 2 Mini46.5%estimated ± 3.3 pp, medium confidence
31Muse Spark Open46.4%estimated ± 3.3 pp, medium confidence
32Grok 4.746.3%measured
33MiMo-V2.6-Flash46.2%estimated ± 4.7 pp, high confidence
34LongCat-2.546.0%estimated ± 4.7 pp, high confidence
35Sakana Fugu-Ultra45.9%estimated ± 3.3 pp, high confidence
36Claude Fable 544.4%estimated ± 2.0 pp, high confidence
37Kimi K343.4%estimated ± 2.0 pp, high confidence
38MiMo-V2.6-Pro43.3%estimated ± 4.7 pp, high confidence
39GPT-6 Astra42.7%estimated ± 2.0 pp, high confidence
40GLM-5.342.6%measured
41GPT-5.1-Codex42.5%estimated ± 4.5 pp, high confidence
42GPT-6.1 Terra42.2%estimated ± 4.9 pp, high confidence
43Ling 3.1 Flash42.2%estimated ± 3.3 pp, high confidence
44Gemini 3.7 Flash42.1%estimated ± 2.0 pp, high confidence
45Gemini 3.8 Flash42.0%estimated ± 2.0 pp, high confidence
46GPT-5.6 Sol41.7%measured
47Muse Spark 1.341.6%measured
48Grok 4.641.4%measured
49GPT-5.6 Terra41.3%measured
50Hy4 flash40.8%estimated ± 4.7 pp, medium confidence
51o340.7%estimated ± 4.5 pp, high confidence
52Inkling 2 Small40.7%estimated ± 2.6 pp, high confidence
53GPT-5.1-Codex-Max40.4%estimated ± 4.5 pp, high confidence
54Beam 1.0 501B-A23B40.2%estimated ± 3.3 pp, high confidence
55Grok 440.0%estimated ± 4.5 pp, high confidence
56Beam 1.0 501B-A23B FP840.0%estimated ± 3.3 pp, high confidence
57GPT-5.539.9%estimated ± 2.0 pp, high confidence
58Hy4 preview39.9%estimated ± 3.3 pp, high confidence
59Beam 1.0 501B-A23B NVFP439.8%estimated ± 3.3 pp, high confidence
60Ornith-1.5-397B39.4%estimated ± 3.3 pp, high confidence
61o4-mini39.0%estimated ± 4.5 pp, high confidence
62Claude Opus 4.838.5%estimated ± 2.0 pp, high confidence
63Claude Opus 4.738.1%estimated ± 2.0 pp, high confidence
64Qwen3.8-Omni-Flash38.1%estimated ± 3.3 pp, high confidence
65GLM-4.637.9%estimated ± 4.5 pp, high confidence
66Grok 4.1 Fast (Reasoning)37.5%estimated ± 4.5 pp, high confidence
67Muse Spark 1.237.5%estimated ± 2.0 pp, high confidence
68Ornith-1.0-397B37.2%estimated ± 3.3 pp, medium confidence
69Muse Spark 1.137.1%estimated ± 2.0 pp, medium confidence
70GLM-5.3-Flash36.8%measured
71Qwen3.8 Max36.7%estimated ± 2.6 pp, medium confidence
72Ripple-65B Preview36.6%estimated ± 3.3 pp, medium confidence
73Grok 4.536.6%estimated ± 2.0 pp, high confidence
74Sage-A1.536.5%estimated ± 3.3 pp, medium confidence
75dots3-note Preview36.3%estimated ± 3.3 pp, medium confidence
76Grok 4 Fast (Reasoning)36.0%estimated ± 4.5 pp, medium confidence
77GPT-5.6 Luna35.9%measured
78Qwen3 Max35.3%estimated ± 4.5 pp, medium confidence
79Atria Dawn Preview35.3%estimated ± 3.3 pp, medium confidence
80Ornith-1.5-35B-A3B35.3%estimated ± 3.3 pp, medium confidence
81Qwen3.8-Flash-Next35.2%estimated ± 2.0 pp, medium confidence
82Laguna S 2.135.1%estimated ± 3.3 pp, medium confidence
83Hy334.9%estimated ± 4.9 pp, medium confidence
84Hy3 Preview34.9%estimated ± 4.9 pp, medium confidence
85MiniMax M334.8%estimated ± 3.3 pp, medium confidence
86Sakana Fugu34.8%estimated ± 3.3 pp, medium confidence
87GPT-5.434.8%estimated ± 2.0 pp, medium confidence
88Qwen3.8 Max Preview34.4%estimated ± 2.0 pp, medium confidence
89Quasar 438B34.4%estimated ± 4.9 pp, medium confidence
90GLM-5.134.4%estimated ± 3.3 pp, medium confidence
91Solar Pro 434.3%estimated ± 4.9 pp, medium confidence
92Claude Sonnet 534.1%measured
93Gemini 3.1 Pro34.1%estimated ± 2.0 pp, medium confidence
94Qwen3.7 Plus33.8%estimated ± 3.3 pp, medium confidence
95Qwen 3.6 Max (preview)33.5%estimated ± 3.3 pp, medium confidence
96MiMo-V2.5-Pro33.5%estimated ± 3.3 pp, medium confidence
97Claude Opus 4.533.4%estimated ± 3.3 pp, medium confidence
98Gemini 3.5 Flash33.3%estimated ± 2.0 pp, medium confidence
99GPT-5.3 Codex33.2%estimated ± 3.3 pp, medium confidence
100Ling 3.0 Flash33.0%estimated ± 3.3 pp, medium confidence
101Qwen3.6 Plus33.0%estimated ± 3.3 pp, medium confidence
102Step 3.7 Flash32.8%estimated ± 3.3 pp, medium confidence
103MiniMax M2.732.7%estimated ± 3.3 pp, medium confidence
104MiMo-V2.532.6%estimated ± 3.3 pp, medium confidence
105Gemini 2.5 Pro32.6%estimated ± 4.9 pp, medium confidence
106Inkling-Small32.5%estimated ± 3.3 pp, medium confidence
107GPT-5.232.3%estimated ± 3.3 pp, medium confidence
108Gemini 3.6 Flash31.9%estimated ± 2.0 pp, medium confidence
109GLM-531.9%estimated ± 3.3 pp, medium confidence
110Apodex 1.131.8%estimated ± 4.9 pp, medium confidence
111Apodex 1.1 Mini31.8%estimated ± 4.9 pp, medium confidence
112Gemma 4 31B31.8%estimated ± 4.9 pp, medium confidence
113Inkling31.7%estimated ± 2.6 pp, medium confidence
114Spark-X2.6-4B31.3%estimated ± 3.3 pp, medium confidence
115Gemini 3.5 Flash-Lite31.2%estimated ± 3.3 pp, medium confidence
116Claude Opus 4.6 (Adaptive)30.9%estimated ± 3.4 pp, medium confidence
117Claude Sonnet 4.630.9%estimated ± 3.4 pp, medium confidence
118Claude Opus 4.5 Thinking30.9%estimated ± 3.4 pp, medium confidence
119Gemini 3 Pro30.9%estimated ± 3.4 pp, medium confidence
120Gemini 3 Flash30.8%estimated ± 3.4 pp, medium confidence
121GPT-5.4 mini30.8%estimated ± 3.4 pp, medium confidence
122GPT-5.2-Codex30.8%estimated ± 3.4 pp, medium confidence
123Grok 4.330.8%estimated ± 3.4 pp, medium confidence
124Claude Sonnet 4.5 Thinking30.8%estimated ± 3.4 pp, medium confidence
125GPT-5.130.8%estimated ± 3.4 pp, medium confidence
126GPT-5.4 nano30.8%estimated ± 3.4 pp, medium confidence
127GLM-4.730.8%estimated ± 3.4 pp, medium confidence
128GPT-530.8%estimated ± 3.4 pp, medium confidence
129Nemotron 3 Ultra30.8%estimated ± 3.4 pp, medium confidence
130Claude Haiku 4.530.8%estimated ± 3.4 pp, medium confidence
131Mistral Medium 3.5 128B30.8%estimated ± 3.4 pp, medium confidence
132Qwen3.5 Flash30.7%estimated ± 3.4 pp, medium confidence
133Gemini 3.1 Flash-Lite30.7%estimated ± 3.4 pp, medium confidence
134Qwen3.6-27B30.7%estimated ± 3.3 pp, medium confidence
135Claude Opus 4.630.6%estimated ± 3.3 pp, medium confidence
136Ling 3.0 Flash VL30.5%estimated ± 4.9 pp, medium confidence
137GLM-5.230.3%estimated ± 2.0 pp, medium confidence
138DeepSeek V4 Pro30.3%estimated ± 2.0 pp, medium confidence
139DeepSeek V4 Flash 073130.2%estimated ± 2.0 pp, medium confidence
140MAI-Thinking-130.1%estimated ± 3.3 pp, medium confidence
141Muse Spark29.8%estimated ± 3.3 pp, medium confidence
142o3-mini29.8%estimated ± 4.5 pp, medium confidence
143Grok 4.2029.4%estimated ± 3.3 pp, medium confidence
144Nemotron 3.5 Super VL 120B A12B29.4%estimated ± 3.3 pp, medium confidence
145Nemotron 3.5 Super VL 120B A12B NVFP429.0%estimated ± 3.3 pp, medium confidence
146Muse Glimmer 30B28.9%estimated ± 3.3 pp, medium confidence
147Qwen3.5 397B28.7%estimated ± 3.3 pp, medium confidence
148Kimi K2.528.6%estimated ± 3.3 pp, medium confidence
149K-EXAONE 2.028.4%estimated ± 4.9 pp, medium confidence
150Ornith-1.0-35B28.3%estimated ± 3.3 pp, medium confidence
151Composer 2.527.7%measured
152Qwen3.6-35B-A3B27.6%estimated ± 3.3 pp, medium confidence
153A.X K227.5%estimated ± 4.9 pp, medium confidence
154Laguna M.127.4%estimated ± 3.3 pp, medium confidence
155Qwen3.8-27B27.2%estimated ± 2.0 pp, medium confidence
156Trinity-Large-Preview27.1%estimated ± 4.9 pp, medium confidence
157Trinity-Large-Thinking27.1%estimated ± 4.9 pp, medium confidence
158Gemma 4 26B A4B26.6%estimated ± 4.9 pp, medium confidence
159Nemotron 3.5 Bolt 74B-A9B26.3%estimated ± 4.9 pp, medium confidence
160Qwen3.5-122B-A10B26.3%estimated ± 4.9 pp, medium confidence
161Laguna XS 2.126.2%estimated ± 3.3 pp, medium confidence
162Ornith-1.5-9B26.1%estimated ± 3.3 pp, medium confidence
163Qwen3.7 Max26.1%estimated ± 2.0 pp, medium confidence
164DeepSeek V3 032425.7%estimated ± 4.9 pp, medium confidence
165GPT-OSS 20B25.6%estimated ± 4.9 pp, medium confidence
166Mistral Small 425.5%estimated ± 4.9 pp, medium confidence
167North Mini Code25.5%estimated ± 4.9 pp, medium confidence
168Laguna XS.225.2%estimated ± 3.3 pp, medium confidence
169Command A+25.2%estimated ± 4.9 pp, medium confidence
170Mercury 2.525.2%estimated ± 4.9 pp, medium confidence
171Mistral Large 323.6%estimated ± 4.9 pp, medium confidence
172Ornith-1.0-9B22.7%estimated ± 3.3 pp, medium confidence
173Kimi K2.622.0%estimated ± 2.0 pp, medium confidence
174LongCat-Flash-Lite-Sparse21.0%estimated ± 3.3 pp, medium confidence
175Kimi K2.7 Code20.6%estimated ± 3.0 pp, medium confidence
176Granite 4.2 30B15.4%estimated ± 3.3 pp, medium confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General