benchgap
Coding

Vibe Code Bench leaderboard

As of 2026-11-21, the highest measured score on Vibe Code Bench is 91.9% by Gemini 4 Argon. 164 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Claude Fable 5.1100.0%estimated ± 10.0 pp, low confidence
2Gemini 4 Argon91.9%measured
3Claude Opus 5.581.7%estimated ± 5.6 pp, low confidence
4GPT-5.6 Sol81.6%estimated ± 10.0 pp, low confidence
5Claude Fable 581.0%estimated ± 5.6 pp, low confidence
6Claude Opus 580.9%estimated ± 5.6 pp, low confidence
7GPT-6 Astra80.8%estimated ± 5.6 pp, low confidence
8Grok 4.678.6%estimated ± 10.0 pp, low confidence
9Mistral Large 478.4%measured
10SWE-278.2%estimated ± 5.6 pp, low confidence
11GPT-5.6 Terra77.9%estimated ± 10.0 pp, low confidence
12Gemini 3.8 Flash76.2%estimated ± 10.0 pp, low confidence
13Kimi K375.9%estimated ± 10.0 pp, low confidence
14Claude Opus 4.875.2%estimated ± 5.6 pp, low confidence
15Claude Haiku 5.575.2%estimated ± 5.6 pp, low confidence
16Claude Sonnet 5.575.0%estimated ± 5.6 pp, low confidence
17SWE-2 Flash74.4%estimated ± 5.6 pp, low confidence
18Muse Spark 1.373.9%estimated ± 10.0 pp, low confidence
19Gemini 3.7 Flash72.6%estimated ± 5.6 pp, low confidence
20Claude Sonnet 571.8%estimated ± 5.6 pp, low confidence
21SWE-1.771.4%estimated ± 5.6 pp, low confidence
22Claude Opus 4.771.0%measured
23GPT-5.569.8%measured
24GLM-5.369.6%estimated ± 10.0 pp, low confidence
25GPT-5.467.4%measured
26Composer 266.5%estimated ± 6.5 pp, low confidence
27Composer 2 Fast66.5%estimated ± 6.5 pp, low confidence
28Claude Mythos 566.3%estimated ± 8.4 pp, low confidence
29Ember-166.3%estimated ± 8.4 pp, low confidence
30Ornith-1.5-397B66.2%estimated ± 8.4 pp, medium confidence
31Qwen3.8-Flash-Next63.2%estimated ± 10.0 pp, low confidence
32Ornith-1.0-397B62.8%estimated ± 8.4 pp, medium confidence
33Inkling 2 Small62.2%estimated ± 8.4 pp, medium confidence
34GPT-5.3 Codex61.8%measured
35Grok 4.561.2%estimated ± 10.0 pp, low confidence
36Muse Spark 1.260.4%estimated ± 10.0 pp, low confidence
37Beam 1.0 501B-A23B59.8%estimated ± 8.4 pp, medium confidence
38Qwen3.8 Max Preview59.1%estimated ± 10.0 pp, low confidence
39GPT-5.6 Luna58.0%estimated ± 10.0 pp, low confidence
40Muse Spark 1.157.6%estimated ± 10.0 pp, low confidence
41Claude Opus 4.657.6%measured
42Claude Opus 4.555.5%estimated ± 8.4 pp, medium confidence
43Beam 1.0 501B-A23B NVFP453.9%estimated ± 8.4 pp, medium confidence
44GPT-5.253.5%measured
45Claude Opus 4.6 (Adaptive)53.5%measured
46MiniMax M352.2%estimated ± 8.4 pp, medium confidence
47Gemini 3.6 Flash51.8%estimated ± 10.0 pp, low confidence
48Claude Sonnet 4.651.5%measured
49Qwen3.7 Max51.2%estimated ± 8.4 pp, medium confidence
50GLM-5.250.5%estimated ± 10.0 pp, low confidence
51DeepSeek V4 Pro49.9%measured
52Inkling-Small49.2%estimated ± 8.4 pp, medium confidence
53Beam 1.0 501B-A23B FP849.2%estimated ± 8.4 pp, medium confidence
54Qwen3.8-27B48.9%estimated ± 10.0 pp, low confidence
55Gemini 3.5 Flash48.7%measured
56GPT-5.4 mini48.0%measured
57GPT-5.2-Codex37.9%measured
58Kimi K2.637.9%measured
59Inkling 2 Mini36.9%estimated ± 8.4 pp, medium confidence
60Quasar 438B35.5%estimated ± 10.0 pp, low confidence
61Apodex 1.1 Mini34.9%estimated ± 10.0 pp, low confidence
62Kimi K2.7 Code34.9%estimated ± 10.0 pp, low confidence
63DeepSeek V4 Flash 073134.1%estimated ± 8.4 pp, medium confidence
64Ornith-1.5-35B-A3B34.1%estimated ± 8.4 pp, medium confidence
65MiMo-V2.5-Pro34.0%estimated ± 10.0 pp, low confidence
66Gemini 3.1 Pro32.0%measured
67Hy332.0%estimated ± 10.0 pp, low confidence
68GLM-5.131.5%measured
69Kimi K2.527.9%estimated ± 6.5 pp, low confidence
70Gemma 4 31B27.7%estimated ± 6.5 pp, low confidence
71GLM-527.7%estimated ± 6.5 pp, low confidence
72Grok 427.7%estimated ± 6.5 pp, low confidence
73GPT-OSS 120B27.7%estimated ± 6.5 pp, low confidence
74DeepSeek V3.227.7%estimated ± 6.5 pp, low confidence
75GPT-OSS 20B27.7%estimated ± 6.5 pp, low confidence
76MiniMax M2.727.0%measured
77GPT-5.4 nano26.1%measured
78BTL-426.0%estimated ± 8.4 pp, medium confidence
79dots3-note Preview26.0%estimated ± 8.4 pp, medium confidence
80Qwen3.6 Plus25.6%measured
81GPT-5.124.6%measured
82Ling 3.0 Flash22.8%estimated ± 10.0 pp, low confidence
83Claude Sonnet 4.5 Thinking22.6%measured
84GPT-5.1-Codex-Max22.2%measured
85Gemini 3.5 Flash-Lite21.5%estimated ± 10.0 pp, low confidence
86MiMo-V2-Pro21.0%estimated ± 8.4 pp, medium confidence
87Claude Opus 4.5 Thinking20.6%measured
88GLM-4.720.6%estimated ± 6.6 pp, low confidence
89o4-mini20.2%estimated ± 6.6 pp, low confidence
90Gemini 3 Flash20.2%measured
91GPT-520.1%measured
92Muse Spark19.7%measured
93Apodex 1.117.6%estimated ± 8.4 pp, medium confidence
94Qwen3.7 Plus17.6%estimated ± 8.4 pp, medium confidence
95Inkling16.6%estimated ± 8.4 pp, medium confidence
96Mistral Medium 3.5 128B16.6%estimated ± 8.4 pp, medium confidence
97Grok 4.316.1%estimated ± 10.0 pp, low confidence
98Qwen3.5 Plus15.7%measured
99MiniMax M2.514.9%measured
100o114.5%estimated ± 10.0 pp, low confidence
101Step 3.7 Flash14.4%estimated ± 10.0 pp, low confidence
102Gemini 3 Pro14.3%measured
103Gemma 4 26B A4B14.2%estimated ± 10.0 pp, low confidence
104GPT-5 mini14.2%measured
105GPT-5.1-Codex13.1%measured
106Claude Sonnet 4.512.8%estimated ± 8.4 pp, medium confidence
107Qwen3.6-27B12.8%estimated ± 8.4 pp, medium confidence
108Claude Haiku 4.5 Thinking11.4%measured
109o1-preview11.3%estimated ± 10.0 pp, low confidence
110Gemma 4 12B9.8%estimated ± 10.0 pp, low confidence
111Command A+8.4%estimated ± 10.0 pp, low confidence
112Mistral Small 47.9%estimated ± 10.0 pp, low confidence
113Ripple-65B Preview7.8%estimated ± 8.4 pp, medium confidence
114Trinity-Large-Preview7.6%estimated ± 10.0 pp, low confidence
115Trinity-Large-Thinking7.6%estimated ± 10.0 pp, low confidence
116Ling 2.6 Flash7.3%estimated ± 10.0 pp, low confidence
117Gemini 1.5 Pro6.7%estimated ± 10.0 pp, low confidence
118Qwen3.5 397B6.3%estimated ± 8.4 pp, medium confidence
119GPT-4 Turbo5.9%estimated ± 10.0 pp, low confidence
120GPT-4.1 mini5.5%estimated ± 10.0 pp, low confidence
121Mistral Large 35.4%estimated ± 10.0 pp, low confidence
122Muse Glimmer 30B5.4%estimated ± 8.4 pp, medium confidence
123Claude 3 Opus5.3%estimated ± 10.0 pp, low confidence
124Llama 4 Maverick4.2%estimated ± 10.0 pp, low confidence
125Grok 4.204.1%measured
126Ornith-1.0-35B4.0%estimated ± 8.4 pp, medium confidence
127Celeris-13.6%estimated ± 10.0 pp, low confidence
128Nemotron 3 Nano 30B3.6%estimated ± 10.0 pp, low confidence
129Qwen3 Max3.5%measured
130Nemotron 3 Nano Omni 30B A3B3.4%estimated ± 10.0 pp, low confidence
131GLM-4.63.1%measured
132Ultravox v0.6 Llama 3.3 70B2.9%estimated ± 10.0 pp, low confidence
133GPT-4o mini2.8%estimated ± 10.0 pp, low confidence
134GPT-4.1 nano2.7%estimated ± 10.0 pp, low confidence
135Gemma 3 27B2.4%estimated ± 10.0 pp, low confidence
136Gemma 4 E4B2.2%estimated ± 10.0 pp, low confidence
137MiMo-V2-Omni2.1%estimated ± 8.4 pp, medium confidence
138Llama 4 Scout1.9%estimated ± 10.0 pp, low confidence
139LFM2.5-2.6B1.8%estimated ± 10.0 pp, low confidence
140Laguna M.11.8%estimated ± 8.4 pp, medium confidence
141Gemma 4 E2B1.7%estimated ± 10.0 pp, low confidence
142Claude 4.1 Opus1.6%estimated ± 8.4 pp, medium confidence
143Hy3 Preview1.5%estimated ± 8.4 pp, medium confidence
144Step 3.5 Flash1.5%estimated ± 8.4 pp, medium confidence
145Flint1.3%estimated ± 8.4 pp, medium confidence
146Grok 4.1 Fast (Reasoning)1.2%measured
147MAI-Thinking-10.7%estimated ± 8.4 pp, medium confidence
148MiMo-V2-Flash0.7%estimated ± 8.4 pp, medium confidence
149Qwen3.6-35B-A3B0.7%estimated ± 8.4 pp, medium confidence
150Claude Haiku 4.50.6%estimated ± 8.4 pp, medium confidence
151Jais-3-MoVA-36B-A4B0.4%estimated ± 8.4 pp, medium confidence
152Gemini 2.5 Pro0.4%measured
153Claude 4 Sonnet0.4%estimated ± 8.4 pp, medium confidence
154MAI-Code-1.1-Flash0.3%estimated ± 8.4 pp, medium confidence
155Claude 4 Opus0.3%estimated ± 8.4 pp, medium confidence
156Jais-3-32B0.3%estimated ± 8.4 pp, medium confidence
157Qwen3.5-27B0.3%estimated ± 8.4 pp, medium confidence
158Qwen3.5-122B-A10B0.2%estimated ± 8.4 pp, medium confidence
159Nemotron 3 Ultra0.2%estimated ± 8.4 pp, medium confidence
160Nemotron 3.5 Super VL 120B A12B0.1%estimated ± 8.4 pp, medium confidence
161Nemotron 3.5 Super VL 120B A12B NVFP40.1%estimated ± 8.4 pp, medium confidence
162Laguna XS 2.10.1%estimated ± 8.4 pp, medium confidence
163Grok Code Fast 10.1%estimated ± 8.4 pp, medium confidence
164Ornith-1.5-9B0.1%estimated ± 8.4 pp, medium confidence
165Solar Pro 40.1%estimated ± 8.4 pp, medium confidence
166Solar Open 20.1%estimated ± 8.4 pp, medium confidence
167Jais-3-375B-A23B0.0%estimated ± 8.4 pp, medium confidence
168Jais-3-7B0.0%estimated ± 8.4 pp, medium confidence
169o30.0%estimated ± 6.6 pp, low confidence
170Laguna XS.20.0%estimated ± 8.4 pp, medium confidence
171Qwen3-Coder 480B-A35B0.0%estimated ± 8.4 pp, medium confidence
172Ornith-1.0-9B0.0%estimated ± 8.4 pp, medium confidence
173Qwen3.5-35B-A3B0.0%estimated ± 8.4 pp, medium confidence
174K-EXAONE 2.00.0%estimated ± 8.4 pp, medium confidence
175LongCat-Flash-Lite-Sparse0.0%estimated ± 8.4 pp, medium confidence
176DeepSeek V3.10.0%estimated ± 8.4 pp, medium confidence
177Kimi K20.0%estimated ± 8.4 pp, medium confidence
178Spark-X2.6-4B0.0%estimated ± 8.4 pp, low confidence
179Claude 3.7 Sonnet0.0%estimated ± 8.4 pp, low confidence
180Ternary Bonsai 2 27B0.0%estimated ± 8.4 pp, low confidence
181Nemotron 3 Super 100B0.0%estimated ± 8.4 pp, low confidence
182GLM-4.7-Flash0.0%estimated ± 8.4 pp, low confidence
183DeepSeek-R1-05280.0%estimated ± 8.4 pp, low confidence
184Granite 4.2 30B0.0%estimated ± 8.4 pp, low confidence
185MiniMax M1 80k0.0%estimated ± 8.4 pp, low confidence
186GPT-4.10.0%estimated ± 8.4 pp, low confidence
187ZAYA1-74B-Preview0.0%estimated ± 8.4 pp, low confidence
188Nemotron 3.5 Lightning 30B A3B NVFP40.0%estimated ± 8.4 pp, low confidence
189LFM2.5-24B-A2B0.0%estimated ± 8.4 pp, low confidence
190K-Exaone0.0%estimated ± 8.4 pp, low confidence
191o3-mini0.0%estimated ± 8.4 pp, low confidence
192LLaDA2.2-flash0.0%estimated ± 8.4 pp, low confidence
193DeepSeek-R10.0%estimated ± 8.4 pp, low confidence
194Claude 3.5 Sonnet0.0%estimated ± 8.4 pp, low confidence
195Granite 4.2 8B0.0%estimated ± 8.4 pp, low confidence
196Mellum2.1-12B-A2.5B-Thinking0.0%estimated ± 8.4 pp, low confidence
197Devstral Small0.0%estimated ± 8.4 pp, low confidence
198MiniCPM5-2B0.0%estimated ± 8.4 pp, low confidence
199Sarvam 105B0.0%estimated ± 8.4 pp, low confidence
200LFM3-1.7B0.0%estimated ± 8.4 pp, low confidence
201DeepSeek V30.0%estimated ± 8.4 pp, low confidence
202Claude 3.5 Haiku0.0%estimated ± 8.4 pp, low confidence
203Gemini 3.1 Flash-Lite0.0%measured
204Grok 4 Fast (Reasoning)0.0%measured

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General