Coding

CursorBench 3.2 leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, the highest measured score on CursorBench 3.2 is 73.4% by Claude Fable 5.1. 160 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Fable 5.1	73.4%	measured
2	Gemini 4 Argon	72.4%	estimated ± 4.1 pp, low confidence
3	Claude Mythos 5	71.4%	estimated ± 3.6 pp, high confidence
4	Grok 4.6	70.8%	measured
5	Grok 4.7	70.6%	estimated ± 2.2 pp, high confidence
6	Claude Fable 5	70.5%	measured
7	Claude Opus 5.5	70.5%	estimated ± 1.6 pp, low confidence
8	Claude Opus 5	70.0%	measured
9	GPT-6 Astra	69.5%	estimated ± 1.6 pp, medium confidence
10	Gemini 3.8 Flash	69.2%	measured
11	Sakana Fugu-Ultra	69.1%	estimated ± 3.6 pp, high confidence
12	Ember-1	68.2%	estimated ± 6.4 pp, low confidence
13	Pareto 26.9	68.1%	estimated ± 6.4 pp, low confidence
14	DeepSeek V4 Flash 0731	67.8%	estimated ± 3.3 pp, medium confidence
15	Muse Spark 1.3	67.6%	estimated ± 2.2 pp, high confidence
16	Pareto 26.10 Preview	67.6%	estimated ± 6.4 pp, low confidence
17	GPT-5.6 Sol	67.2%	measured
18	Muse Spark 1.2	67.1%	estimated ± 3.3 pp, medium confidence
19	GLM-5.3-Flash	67.0%	estimated ± 4.1 pp, low confidence
20	Grok 4.5	66.7%	measured
21	SWE-2	66.7%	estimated ± 1.6 pp, medium confidence
22	GPT-6 Sol	65.8%	estimated ± 5.3 pp, medium confidence
23	Step 5 Preview	65.3%	estimated ± 3.6 pp, medium confidence
24	Qwen3.8-27B	64.9%	estimated ± 3.3 pp, medium confidence
25	GPT-5.6 Terra	64.9%	measured
26	Qwen3.8 Max	64.2%	estimated ± 3.3 pp, medium confidence
27	Hy4 preview	64.2%	estimated ± 3.6 pp, high confidence
28	Beam	64.0%	estimated ± 3.6 pp, high confidence
29	Ornith-1.5-397B	63.7%	estimated ± 3.6 pp, high confidence
30	Claude Haiku 5.5	63.4%	estimated ± 1.6 pp, medium confidence
31	Claude Sonnet 5.5	63.2%	estimated ± 1.6 pp, medium confidence
32	Claude Opus 4.7 (Adaptive)	62.9%	estimated ± 3.6 pp, high confidence
33	GLM-5.3	62.7%	estimated ± 3.3 pp, medium confidence
34	Claude Opus 4.8	62.3%	measured
35	Qwen3.8-Omni-Flash	62.0%	estimated ± 3.6 pp, high confidence
36	Claude Haiku 4.5	61.5%	estimated ± 3.3 pp, medium confidence
37	Claude Sonnet 5	61.5%	measured
38	Qwen3.8-Flash-Next	61.1%	estimated ± 3.6 pp, high confidence
39	GPT-5.6 Luna	61.1%	measured
40	GPT-6 Luna	61.0%	estimated ± 5.3 pp, medium confidence
41	Kimi K3	60.8%	measured
42	Ornith-1.0-397B	60.8%	estimated ± 3.6 pp, high confidence
43	Gemini 3.7 Flash	60.7%	estimated ± 1.6 pp, medium confidence
44	GPT-6.1 Sol	60.4%	estimated ± 5.3 pp, medium confidence
45	Mistral Large 4	60.4%	estimated ± 5.3 pp, medium confidence
46	Ling 3.1 Flash	60.2%	estimated ± 5.3 pp, medium confidence
47	Muse Spark 1.1	60.0%	estimated ± 3.6 pp, high confidence
48	Qwen3.8 Max Preview	59.7%	estimated ± 4.4 pp, high confidence
49	SWE-1.7	59.4%	estimated ± 1.6 pp, low confidence
50	dots3-note Preview	59.4%	estimated ± 3.6 pp, high confidence
51	GPT-5.1-Codex	59.1%	estimated ± 7.2 pp, low confidence
52	GPT-5.1-Codex-Max	59.1%	estimated ± 7.2 pp, low confidence
53	GLM-4.5	59.1%	estimated ± 7.2 pp, low confidence
54	GLM-4.6	59.1%	estimated ± 7.2 pp, low confidence
55	Grok Code Fast 1	59.1%	estimated ± 7.2 pp, low confidence
56	Qwen3.7 Max	58.9%	estimated ± 3.6 pp, high confidence
57	GPT-5.5	58.4%	measured
58	Atria Dawn Preview	57.6%	estimated ± 3.6 pp, high confidence
59	Ornith-1.5-35B-A3B	57.6%	estimated ± 3.6 pp, high confidence
60	Laguna S 2.1	57.3%	estimated ± 3.6 pp, high confidence
61	MiniMax M3	56.8%	estimated ± 3.6 pp, high confidence
62	Sakana Fugu	56.8%	estimated ± 3.6 pp, high confidence
63	Kimi K2.6	56.2%	estimated ± 3.6 pp, high confidence
64	Composer 2.5	56.1%	measured
65	GLM-5.1	55.9%	estimated ± 3.6 pp, high confidence
66	Gemini 3.1 Pro	55.8%	estimated ± 4.4 pp, high confidence
67	Claude Opus 4.7	55.6%	estimated ± 1.6 pp, low confidence
68	Gemini 3 Flash	55.5%	estimated ± 5.4 pp, low confidence
69	GLM-5.2	55.0%	measured
70	GPT-5.4	54.9%	estimated ± 3.6 pp, high confidence
71	Qwen3.7 Plus	54.7%	estimated ± 3.6 pp, high confidence
72	Qwen 3.6 Max (preview)	54.3%	estimated ± 3.6 pp, high confidence
73	MiMo-V2.5-Pro	54.1%	estimated ± 3.6 pp, high confidence
74	GPT-5.2-Codex	54.1%	estimated ± 5.4 pp, low confidence
75	Claude Opus 4.5	53.9%	estimated ± 3.6 pp, high confidence
76	Gemini 3.6 Flash	53.5%	measured
77	GPT-5.3 Codex	53.5%	estimated ± 3.6 pp, high confidence
78	Ling 3.0 Flash	53.1%	estimated ± 3.6 pp, high confidence
79	Qwen3.6 Plus	53.1%	estimated ± 3.6 pp, high confidence
80	Step 3.7 Flash	52.6%	estimated ± 3.6 pp, high confidence
81	MiniMax M2.7	52.5%	estimated ± 3.6 pp, high confidence
82	MiMo-V2.5	52.3%	estimated ± 3.6 pp, high confidence
83	Inkling-Small	52.0%	estimated ± 3.6 pp, high confidence
84	GPT-5.2	51.4%	estimated ± 3.6 pp, high confidence
85	DeepSeek V4 Pro 0813	51.1%	estimated ± 3.6 pp, high confidence
86	GLM-5	50.6%	estimated ± 3.6 pp, high confidence
87	Kimi K2.7 Code	49.7%	measured
88	Qwen3.5 Flash	49.6%	estimated ± 5.4 pp, low confidence
89	Inkling	49.1%	estimated ± 3.6 pp, medium confidence
90	Gemini 3.5 Flash-Lite	48.9%	estimated ± 3.6 pp, medium confidence
91	Gemini 3.5 Flash	48.8%	measured
92	Gemini 3.1 Flash-Lite	48.6%	estimated ± 5.4 pp, low confidence
93	Qwen3.6-27B	47.6%	estimated ± 3.6 pp, medium confidence
94	Quasar 438B	46.5%	estimated ± 4.4 pp, high confidence
95	MAI-Thinking-1	46.2%	estimated ± 3.6 pp, medium confidence
96	Apodex 1.1	46.0%	estimated ± 4.4 pp, high confidence
97	Apodex 1.1 Mini	46.0%	estimated ± 4.4 pp, high confidence
98	Muse Spark	45.4%	estimated ± 3.6 pp, medium confidence
99	Solar Pro 4	45.1%	estimated ± 5.3 pp, low confidence
100	Ling 3.0 Flash VL	44.4%	estimated ± 5.3 pp, low confidence
101	Grok 4.20	44.1%	estimated ± 3.6 pp, medium confidence
102	Hy3	43.9%	estimated ± 4.4 pp, medium confidence
103	Hy3 Preview	43.9%	estimated ± 4.4 pp, medium confidence
104	Muse Glimmer 30B	42.8%	estimated ± 3.6 pp, medium confidence
105	GPT-5.4 mini	42.5%	estimated ± 1.6 pp, low confidence
106	Claude Opus 4.6	42.4%	estimated ± 1.6 pp, low confidence
107	Qwen3.5 397B	42.2%	estimated ± 3.6 pp, medium confidence
108	Kimi K2.5	41.8%	estimated ± 3.6 pp, medium confidence
109	Ornith-1.0-35B	41.1%	estimated ± 3.6 pp, medium confidence
110	GPT-5.4 nano	41.0%	estimated ± 4.4 pp, medium confidence
111	K-EXAONE 2.0	40.9%	estimated ± 5.3 pp, low confidence
112	A.X K2	39.3%	estimated ± 5.3 pp, low confidence
113	Claude Sonnet 4.6	39.1%	estimated ± 1.6 pp, low confidence
114	Qwen3.6-35B-A3B	39.1%	estimated ± 3.6 pp, medium confidence
115	Laguna M.1	38.4%	estimated ± 3.6 pp, medium confidence
116	DeepSeek V3 0324	36.1%	estimated ± 5.3 pp, low confidence
117	North Mini Code	35.8%	estimated ± 5.3 pp, low confidence
118	Ling 3.0 Flash FP8	35.5%	estimated ± 4.4 pp, medium confidence
119	Mercury 2.5	35.3%	estimated ± 5.3 pp, low confidence
120	MiMo-V2-Flash	34.7%	estimated ± 4.4 pp, medium confidence
121	Laguna XS 2.1	34.7%	estimated ± 3.6 pp, medium confidence
122	Ornith-1.5-9B	34.5%	estimated ± 3.6 pp, medium confidence
123	GPT-5.1	34.3%	estimated ± 4.4 pp, medium confidence
124	Nemotron 3 Ultra	34.2%	estimated ± 4.4 pp, medium confidence
125	Mistral Medium 3.5 128B	32.0%	estimated ± 4.4 pp, medium confidence
126	Kimi K2.5 (Reasoning)	31.9%	estimated ± 4.4 pp, medium confidence
127	Laguna XS.2	31.6%	estimated ± 3.6 pp, medium confidence
128	Qwen3.5-122B-A10B	30.9%	estimated ± 4.4 pp, medium confidence
129	GLM-4.7	30.5%	estimated ± 4.4 pp, medium confidence
130	Gemma 4 31B	28.9%	estimated ± 4.4 pp, medium confidence
131	Grok 4.3	27.9%	estimated ± 4.4 pp, medium confidence
132	MiMo-V2.6-Pro	27.5%	estimated ± 3.6 pp, low confidence
133	MiMo-V2.6-Flash	27.1%	estimated ± 3.6 pp, low confidence
134	o1	25.8%	estimated ± 4.4 pp, medium confidence
135	Gemma 4 26B A4B	25.4%	estimated ± 4.4 pp, medium confidence
136	GPT-5 (high)	24.2%	estimated ± 4.4 pp, medium confidence
137	Nemotron 3 Super 100B	24.1%	estimated ± 4.4 pp, medium confidence
138	Ornith-1.0-9B	23.7%	estimated ± 3.6 pp, medium confidence
139	DeepSeek V4.1 Flash	21.8%	estimated ± 3.6 pp, low confidence
140	o1-preview	21.3%	estimated ± 4.4 pp, medium confidence
141	Gemini 2.5 Pro	20.7%	estimated ± 4.4 pp, medium confidence
142	K-Exaone	19.8%	estimated ± 4.4 pp, medium confidence
143	Gemma 4 12B	18.9%	estimated ± 4.4 pp, medium confidence
144	LongCat-Flash-Lite-Sparse	18.7%	estimated ± 3.6 pp, medium confidence
145	GPT-OSS 120B	18.6%	estimated ± 4.4 pp, medium confidence
146	Command A+	16.7%	estimated ± 4.4 pp, medium confidence
147	Nemotron 3.5 Lightning 30B A3B NVFP4	15.9%	estimated ± 4.4 pp, medium confidence
148	Mistral Small 4	15.9%	estimated ± 4.4 pp, medium confidence
149	Mistral Small 4 (Reasoning)	15.9%	estimated ± 4.4 pp, medium confidence
150	Trinity-Large-Preview	15.3%	estimated ± 4.4 pp, medium confidence
151	Trinity-Large-Thinking	15.3%	estimated ± 4.4 pp, medium confidence
152	Ling 2.6 Flash	14.9%	estimated ± 4.4 pp, medium confidence
153	Solar Pro 3	14.6%	estimated ± 5.3 pp, low confidence
154	Granite 4.2 3B	14.3%	estimated ± 5.3 pp, low confidence
155	Gemini 1.5 Pro	13.8%	estimated ± 4.4 pp, medium confidence
156	DeepSeek V3	13.4%	estimated ± 4.4 pp, medium confidence
157	Ling 3.0 Tiny	12.5%	estimated ± 5.3 pp, low confidence
158	GPT-4 Turbo	12.4%	estimated ± 4.4 pp, medium confidence
159	GPT-OSS 20B	11.9%	estimated ± 4.4 pp, medium confidence
160	GPT-4.1 mini	11.6%	estimated ± 4.4 pp, medium confidence
161	Mistral Large 3	11.5%	estimated ± 4.4 pp, medium confidence
162	Claude 3 Opus	11.1%	estimated ± 4.4 pp, medium confidence
163	Llama 4 Maverick	9.1%	estimated ± 4.4 pp, medium confidence
164	Celeris-1	8.0%	estimated ± 4.4 pp, medium confidence
165	Nemotron 3 Nano 30B	7.9%	estimated ± 4.4 pp, medium confidence
166	Nemotron 3 Nano Omni 30B A3B	7.6%	estimated ± 4.4 pp, medium confidence
167	Granite 4.2 30B	6.9%	estimated ± 3.6 pp, medium confidence
168	Ultravox v0.6 Llama 3.3 70B	6.5%	estimated ± 4.4 pp, medium confidence
169	GPT-4o mini	6.2%	estimated ± 4.4 pp, medium confidence
170	GPT-4.1 nano	6.0%	estimated ± 4.4 pp, medium confidence
171	Gemma 3 27B	5.4%	estimated ± 4.4 pp, medium confidence
172	Gemma 4 E4B	5.0%	estimated ± 4.4 pp, medium confidence
173	Llama 4 Scout	4.4%	estimated ± 4.4 pp, medium confidence
174	LFM2.5-2.6B	4.1%	estimated ± 4.4 pp, medium confidence
175	LLaDA2.2-flash	3.9%	estimated ± 3.6 pp, medium confidence
176	Gemma 4 E2B	3.8%	estimated ± 4.4 pp, medium confidence
177	Granite 4.2 8B	0.3%	estimated ± 3.6 pp, medium confidence
178	MiniCPM5-2B	0.0%	estimated ± 3.6 pp, medium confidence

CursorBench 3.2 leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General