Coding

CursorBench 3.1 leaderboard

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

As of 2026-10-07, the highest measured score on CursorBench 3.1 is 70.6% by Claude Fable 5. 161 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Opus 5.5	85.7%	estimated ± 3.4 pp, low confidence
2	Claude Fable 5.1	76.3%	estimated ± 3.4 pp, low confidence
3	Gemini 4 Argon	73.1%	estimated ± 3.4 pp, low confidence
4	Claude Sonnet 5.5	71.1%	estimated ± 3.4 pp, medium confidence
5	MiMo-V2.6-Pro	70.9%	estimated ± 3.4 pp, medium confidence
6	Claude Fable 5	70.6%	measured
7	Kimi K3	67.4%	estimated ± 3.4 pp, medium confidence
8	Claude Mythos 5	66.9%	estimated ± 5.6 pp, low confidence
9	GLM-5.3	66.2%	estimated ± 3.4 pp, medium confidence
10	Step 5 Preview	65.9%	estimated ± 3.4 pp, medium confidence
11	Muse Spark 1.1	65.7%	estimated ± 3.4 pp, medium confidence
12	Muse Spark 1.3	65.7%	estimated ± 3.4 pp, medium confidence
13	Gemini 3.1 Pro	65.5%	estimated ± 3.4 pp, medium confidence
14	Composer 2.5	63.2%	measured
15	Sakana Fugu-Ultra	63.2%	estimated ± 5.6 pp, low confidence
16	GPT-6 Sol	62.7%	estimated ± 3.4 pp, medium confidence
17	Grok 4.7	62.2%	estimated ± 3.4 pp, medium confidence
18	Muse Spark 1.2	62.2%	estimated ± 3.4 pp, medium confidence
19	Gemini 3.7 Flash	61.8%	estimated ± 3.4 pp, medium confidence
20	GPT-5.6 Sol	61.5%	estimated ± 3.4 pp, medium confidence
21	Gemini 3.8 Flash	60.3%	estimated ± 3.4 pp, medium confidence
22	GPT-6 Astra	60.0%	estimated ± 3.4 pp, medium confidence
23	Grok 4.6	60.0%	estimated ± 3.4 pp, medium confidence
24	Claude Opus 5	59.8%	estimated ± 3.4 pp, medium confidence
25	Qwen3.8 Max	59.6%	estimated ± 5.6 pp, low confidence
26	GPT-5.5	59.2%	measured
27	Claude Opus 4.8	58.4%	measured
28	Hy4 preview	58.3%	estimated ± 5.6 pp, low confidence
29	Beam	58.2%	estimated ± 5.6 pp, low confidence
30	Ornith-1.5-397B	57.9%	estimated ± 5.6 pp, low confidence
31	Qwen3.8-Omni-Flash	56.8%	estimated ± 5.6 pp, low confidence
32	Claude Haiku 5.5	56.3%	estimated ± 3.4 pp, medium confidence
33	GPT-5.6 Terra	56.3%	estimated ± 3.4 pp, medium confidence
34	Grok 4.5	56.3%	estimated ± 3.4 pp, medium confidence
35	Claude Opus 4.7	56.2%	estimated ± 5.8 pp, low confidence
36	Ornith-1.0-397B	56.1%	estimated ± 5.6 pp, low confidence
37	GPT-6 Luna	55.3%	estimated ± 3.4 pp, medium confidence
38	dots3-note Preview	55.3%	estimated ± 5.6 pp, low confidence
39	Claude Opus 4.7 (Adaptive)	55.2%	estimated ± 4.8 pp, medium confidence
40	Claude Sonnet 5	54.6%	estimated ± 3.4 pp, medium confidence
41	Atria Dawn Preview	54.4%	estimated ± 5.6 pp, low confidence
42	Ornith-1.5-35B-A3B	54.4%	estimated ± 5.6 pp, low confidence
43	GPT-6.1 Sol	54.4%	estimated ± 3.4 pp, medium confidence
44	Mistral Large 4	54.4%	estimated ± 3.4 pp, medium confidence
45	Laguna S 2.1	54.2%	estimated ± 5.6 pp, low confidence
46	Ling 3.1 Flash	54.1%	estimated ± 3.4 pp, medium confidence
47	Sakana Fugu	54.0%	estimated ± 5.6 pp, low confidence
48	GPT-5.6 Luna	52.9%	estimated ± 3.4 pp, medium confidence
49	Qwen 3.6 Max (preview)	52.8%	estimated ± 5.6 pp, low confidence
50	Claude Opus 4.5	52.7%	estimated ± 5.6 pp, low confidence
51	GPT-5.3 Codex	52.5%	estimated ± 5.6 pp, low confidence
52	Gemini 3.6 Flash	52.4%	estimated ± 3.4 pp, medium confidence
53	MiMo-V2.5	52.0%	estimated ± 5.6 pp, low confidence
54	GPT-5.2	51.7%	estimated ± 5.6 pp, low confidence
55	GLM-5	51.3%	estimated ± 5.6 pp, low confidence
56	GPT-5.4	50.2%	estimated ± 4.8 pp, medium confidence
57	Claude Opus 4.6	50.1%	estimated ± 5.6 pp, low confidence
58	Gemini 3.5 Flash	49.8%	measured
59	MAI-Thinking-1	49.7%	estimated ± 5.6 pp, low confidence
60	GPT-5.4 mini	49.2%	estimated ± 3.4 pp, medium confidence
61	Qwen3.8 Max Preview	49.2%	estimated ± 3.4 pp, medium confidence
62	Grok 4.20	49.0%	estimated ± 5.6 pp, low confidence
63	Gemini 3 Flash	49.0%	estimated ± 5.8 pp, low confidence
64	Claude Sonnet 4.6	48.8%	measured
65	DeepSeek V4.1 Flash	48.7%	estimated ± 3.4 pp, medium confidence
66	Qwen3.5 397B	48.4%	estimated ± 5.6 pp, low confidence
67	Ornith-1.0-35B	48.0%	estimated ± 5.6 pp, low confidence
68	GLM-5.3-Flash	47.9%	estimated ± 3.4 pp, medium confidence
69	Muse Spark	47.9%	estimated ± 4.8 pp, low confidence
70	Qwen3.6 Plus	47.9%	estimated ± 4.8 pp, low confidence
71	GPT-5.1	47.9%	estimated ± 4.8 pp, low confidence
72	MiMo-V2-Flash	47.9%	estimated ± 4.8 pp, low confidence
73	Claude 3 Opus	47.9%	estimated ± 4.8 pp, low confidence
74	Gemini 1.5 Pro	47.9%	estimated ± 4.8 pp, low confidence
75	Gemma 4 12B	47.9%	estimated ± 4.8 pp, low confidence
76	Gemma 4 E2B	47.9%	estimated ± 4.8 pp, low confidence
77	GLM-4.7	47.9%	estimated ± 4.8 pp, low confidence
78	GPT-4.1 mini	47.9%	estimated ± 4.8 pp, low confidence
79	GPT-4.1 nano	47.9%	estimated ± 4.8 pp, low confidence
80	GPT-4 Turbo	47.9%	estimated ± 4.8 pp, low confidence
81	GPT-4o mini	47.9%	estimated ± 4.8 pp, low confidence
82	GPT-5 (high)	47.9%	estimated ± 4.8 pp, low confidence
83	K-Exaone	47.9%	estimated ± 4.8 pp, low confidence
84	Kimi K2.5	47.9%	estimated ± 4.8 pp, low confidence
85	Kimi K2.5 (Reasoning)	47.9%	estimated ± 4.8 pp, low confidence
86	Ling 2.6 Flash	47.9%	estimated ± 4.8 pp, low confidence
87	Nemotron 3 Nano Omni 30B A3B	47.9%	estimated ± 4.8 pp, low confidence
88	o1	47.9%	estimated ± 4.8 pp, low confidence
89	o1-preview	47.9%	estimated ± 4.8 pp, low confidence
90	Ultravox v0.6 Llama 3.3 70B	47.9%	estimated ± 4.8 pp, low confidence
91	Kimi K2.6	47.6%	measured
92	MiMo-V2.6-Flash	47.2%	estimated ± 3.4 pp, low confidence
93	Laguna M.1	47.1%	estimated ± 5.6 pp, low confidence
94	GLM-5.2	47.0%	estimated ± 3.4 pp, low confidence
95	DeepSeek V4 Pro 0813	46.5%	estimated ± 3.4 pp, low confidence
96	GPT-5.2-Codex	46.3%	estimated ± 5.8 pp, low confidence
97	Laguna XS 2.1	45.9%	estimated ± 5.6 pp, low confidence
98	Ornith-1.5-9B	45.9%	estimated ± 5.6 pp, low confidence
99	MiMo-V2.5-Pro	45.5%	estimated ± 3.4 pp, low confidence
100	Qwen3.8-Flash-Next	45.5%	estimated ± 3.4 pp, low confidence
101	Laguna XS.2	45.0%	estimated ± 5.6 pp, low confidence
102	DeepSeek V4 Flash 0731	44.7%	estimated ± 3.4 pp, low confidence
103	MiniMax M2.7	44.3%	estimated ± 3.4 pp, low confidence
104	Inkling-Small	43.3%	estimated ± 3.4 pp, low confidence
105	Qwen3.7 Max	42.8%	estimated ± 3.4 pp, low confidence
106	Ornith-1.0-9B	42.4%	estimated ± 5.6 pp, low confidence
107	LongCat-Flash-Lite-Sparse	40.6%	estimated ± 5.6 pp, low confidence
108	Hy3	40.6%	estimated ± 3.4 pp, low confidence
109	Hy3 Preview	40.6%	estimated ± 3.4 pp, low confidence
110	Claude Haiku 4.5	40.4%	estimated ± 5.8 pp, low confidence
111	Grok 4.3	39.8%	estimated ± 3.4 pp, low confidence
112	Quasar 438B	39.3%	estimated ± 3.4 pp, low confidence
113	Kimi K2.7 Code	38.6%	estimated ± 3.4 pp, low confidence
114	Qwen3.5 Flash	38.1%	estimated ± 5.8 pp, low confidence
115	GPT-5.4 nano	37.1%	estimated ± 3.4 pp, low confidence
116	MiniMax M3	36.9%	estimated ± 3.4 pp, low confidence
117	Inkling	36.6%	estimated ± 3.4 pp, low confidence
118	Gemini 3.1 Flash-Lite	36.5%	estimated ± 5.8 pp, low confidence
119	Qwen3.8-27B	35.6%	estimated ± 3.4 pp, low confidence
120	Gemini 2.5 Pro	34.9%	estimated ± 3.4 pp, low confidence
121	Qwen3.7 Plus	34.4%	estimated ± 3.4 pp, low confidence
122	Apodex 1.1	32.9%	estimated ± 3.4 pp, low confidence
123	Apodex 1.1 Mini	32.9%	estimated ± 3.4 pp, low confidence
124	Gemma 4 31B	32.9%	estimated ± 3.4 pp, low confidence
125	LLaDA2.2-flash	31.7%	estimated ± 5.6 pp, low confidence
126	Muse Glimmer 30B	31.4%	estimated ± 3.4 pp, low confidence
127	GLM-5.1	31.2%	estimated ± 3.4 pp, low confidence
128	Solar Pro 4	30.7%	estimated ± 3.4 pp, low confidence
129	Ling 3.0 Flash VL	29.7%	estimated ± 3.4 pp, low confidence
130	Step 3.7 Flash	29.0%	estimated ± 3.4 pp, low confidence
131	Qwen3.6-27B	26.3%	estimated ± 3.4 pp, low confidence
132	K-EXAONE 2.0	24.3%	estimated ± 3.4 pp, low confidence
133	Ling 3.0 Flash	24.3%	estimated ± 3.4 pp, low confidence
134	Ling 3.0 Flash FP8	24.3%	estimated ± 3.4 pp, low confidence
135	Gemini 3.5 Flash-Lite	22.6%	estimated ± 3.4 pp, low confidence
136	A.X K2	21.8%	estimated ± 3.4 pp, low confidence
137	Trinity-Large-Preview	20.8%	estimated ± 3.4 pp, low confidence
138	Trinity-Large-Thinking	20.8%	estimated ± 3.4 pp, low confidence
139	Nemotron 3 Ultra	20.1%	estimated ± 3.4 pp, low confidence
140	Mistral Medium 3.5 128B	19.8%	estimated ± 3.4 pp, low confidence
141	Gemma 4 26B A4B	19.3%	estimated ± 3.4 pp, low confidence
142	Qwen3.5-122B-A10B	18.6%	estimated ± 3.4 pp, low confidence
143	DeepSeek V3 0324	16.9%	estimated ± 3.4 pp, low confidence
144	GPT-OSS 20B	16.6%	estimated ± 3.4 pp, low confidence
145	Mistral Small 4	16.4%	estimated ± 3.4 pp, low confidence
146	Mistral Small 4 (Reasoning)	16.4%	estimated ± 3.4 pp, low confidence
147	North Mini Code	16.4%	estimated ± 3.4 pp, low confidence
148	Command A+	15.6%	estimated ± 3.4 pp, low confidence
149	Mercury 2.5	15.6%	estimated ± 3.4 pp, low confidence
150	Granite 4.2 30B	13.9%	estimated ± 3.4 pp, low confidence
151	Mistral Large 3	11.0%	estimated ± 3.4 pp, low confidence
152	Qwen3.6-35B-A3B	11.0%	estimated ± 3.4 pp, low confidence
153	Nemotron 3 Super 100B	10.0%	estimated ± 3.4 pp, low confidence
154	DeepSeek V3	9.0%	estimated ± 3.4 pp, low confidence
155	GPT-OSS 120B	4.6%	estimated ± 3.4 pp, low confidence
156	Celeris-1	0.0%	estimated ± 3.4 pp, low confidence
157	Gemma 3 27B	0.0%	estimated ± 3.4 pp, low confidence
158	Gemma 4 E4B	0.0%	estimated ± 3.4 pp, low confidence
159	Granite 4.2 3B	0.0%	estimated ± 3.4 pp, low confidence
160	Granite 4.2 8B	0.0%	estimated ± 3.4 pp, low confidence
161	LFM2.5-2.6B	0.0%	estimated ± 3.4 pp, low confidence
162	Ling 3.0 Tiny	0.0%	estimated ± 3.4 pp, low confidence
163	Llama 4 Maverick	0.0%	estimated ± 3.4 pp, low confidence
164	Llama 4 Scout	0.0%	estimated ± 3.4 pp, low confidence
165	MiniCPM5-2B	0.0%	estimated ± 3.4 pp, low confidence
166	Nemotron 3.5 Lightning 30B A3B NVFP4	0.0%	estimated ± 3.4 pp, low confidence
167	Nemotron 3 Nano 30B	0.0%	estimated ± 3.4 pp, low confidence
168	Solar Pro 3	0.0%	estimated ± 3.4 pp, low confidence

CursorBench 3.1 leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General