Skip to main content
BenchLM
Data

Vibe Code Bench v1.1 (Vibe Code Bench)

We show this table for reference; we do not rank on it.

Data verified

Vals.ai benchmark for evaluating whether models can build complete web applications from natural language specifications in a production-like development environment.

Evaluation results published by Vals AI. Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.

Vibe Code score on Vibe Code Bench — October 27, 2026

We mirror the published vibe code score view for Vibe Code Bench. Claude Sonnet 5.5 leads the public snapshot at 92.39%, followed by Gemini 4 Argon (91.91%) and Claude Haiku 5.5 (90.44%). We do not use these results to rank models overall.

1Closed

Claude Sonnet 5.5

Anthropic

OpenHands

OpenHands

92.39%
Overall 81.84Context 1M
2Closed

Gemini 4 Argon

Google

OpenHands · high reasoning

OpenHands

91.91%
Overall 82.28Context 1M
3Closed

Claude Haiku 5.5

Anthropic

OpenHands

OpenHands

90.44%
Overall 66.42Context 1M

121 modelsCodingCurrentDisplay onlyUpdated October 27, 2026

Vibe Code score table (121 models)

Vibe Code score results for Vibe Code Bench
RankModel / configurationScoreParameters (B)Open / closed
1Claude Sonnet 5.5AnthropicOpenHandsOpenHands
92.39%
Not reportedClosed
2Gemini 4 ArgonGoogleOpenHands · high reasoningOpenHands
91.91%
Not reportedClosed
3Claude Haiku 5.5AnthropicOpenHandsOpenHands
90.44%
Not reportedClosed
4Claude Fable 5AnthropicOpenHandsOpenHands
90.35%
Not reportedClosed
5Claude Opus 5.5AnthropicOpenHandsOpenHands
90.29%
Not reportedClosed
6Claude Fable 5.1AnthropicOpenHandsOpenHands
90.26%
Not reportedClosed
7GPT-6 AstraOpenAIOpenHands · max reasoningOpenHands
89.59%
Not reportedClosed
8Gemini 4 CarbonGoogleOpenHands · high reasoningOpenHands
89.06%
Not reportedClosed
9GPT-6.1 SolOpenAIOpenHands · max reasoningOpenHands
88.93%
Not reportedClosed
10Claude Opus 5AnthropicOpenHandsOpenHands
88.40%
Not reportedClosed
11GPT-6 SolOpenAIOpenHands · max reasoningOpenHands
87.82%
Not reportedClosed
12DeepSeek V4.1 Flash 1017DeepSeekOpenHands · max reasoningOpenHands
87.73%
Not reportedOpen
13Grok 4.7xAIOpenHands · xhigh reasoningOpenHands
86.17%
Not reportedClosed
14Muse Wonder FullMetaOpenHands · max reasoningOpenHands
86.00%
Not reportedClosed
15Muse Spark 1.3 MaxMetaOpenHands · max reasoningOpenHands
85.86%
Not reportedUnknown
16MiMo-V2.6-ProXiaomiOpenHandsOpenHands
85.22%
Not reportedOpen
17Kimi K3Moonshot AIOpenHands · max reasoningOpenHands
84.97%
Not reportedOpen
18Kimi K3.1 ProMoonshot AIOpenHands · max reasoningOpenHands
84.89%
Not reportedPending
19DeepSeek V4.1 Flash 1017DeepSeekOpenHands · high reasoningOpenHands
84.74%
Not reportedOpen
20GPT-6.1 TerraOpenAIOpenHands · max reasoningOpenHands
84.00%
Not reportedClosed
21Ember-1FireworksOpenHandsOpenHands
83.87%
Not reportedClosed
22Muse WonderMetaOpenHands · max reasoningOpenHands
83.09%
Not reportedClosed
23Muse Spark 1.3MetaOpenHands · xhigh reasoningOpenHands
82.86%
Not reportedClosed
24Claude Opus 4.8AnthropicOpenHandsOpenHands
82.72%
Not reportedClosed
25DeepSeek V4 Pro 0813DeepSeekOpenHands · max reasoningOpenHands
82.30%
Not reportedOpen
26GPT-6 LunaOpenAIOpenHands · max reasoningOpenHands
81.65%
Not reportedClosed
27Claude Sonnet 5AnthropicOpenHandsOpenHands
81.33%
Not reportedClosed
28GPT-5.6 SolOpenAIOpenHands · max reasoningOpenHands
80.50%
Not reportedClosed
29Muse Spark 1.2MetaOpenHands · xhigh reasoningOpenHands
79.10%
Not reportedClosed
30MiMo-V2.6-FlashXiaomiOpenHandsOpenHands
78.96%
Not reportedOpen
31Gemini 3.8 FlashGoogleOpenHands · high reasoningOpenHands
78.65%
Not reportedClosed
32Mistral Large 4MistralOpenHands · high reasoningOpenHands
78.40%
Not reportedOpen
33GLM-5.3Z.AIOpenHands · max reasoningOpenHands
78.13%
Not reportedOpen
34Claude Opus 4.8 Claude CodeAnthropicClaude CodeClaude Code
77.48%
Not reportedUnknown
35Hy4 previewTencentOpenHandsOpenHands
77.48%
Not reportedOpen
36Qwen3.8 Max InternalAlibabaOpenHandsOpenHands
77.14%
Not reportedUnknown
37GPT-5.6 LunaOpenAIOpenHands · max reasoningOpenHands
77.06%
Not reportedClosed
38Grok 4.6xAIOpenHands · high reasoningOpenHands
76.24%
Not reportedClosed
39Muse Spark OpenMetaOpenHands · xhigh reasoningOpenHands
76.19%
Not reportedOpen
40DeepSeek V4 Flash 0731DeepSeekOpenHands · high reasoningOpenHands
74.74%
Not reportedOpen
41GPT-5.6 TerraOpenAIOpenHands · max reasoningOpenHands
74.59%
Not reportedClosed
42MiniMax M3.1 FlashMiniMaxOpenHandsOpenHands
73.94%
Not reportedClosed
43Muse Spark 1.1MetaOpenHands · xhigh reasoningOpenHands
72.16%
Not reportedClosed
44Claude Opus 4.7AnthropicOpenHandsOpenHands
71.00%
Not reportedClosed
45Step 5 PreviewStepFunOpenHandsOpenHands
70.40%
Not reportedOpen
46Gemini 3.7 FlashGoogleOpenHands · high reasoningOpenHands
70.39%
Not reportedClosed
47GPT-5.5OpenAIOpenHands · xhigh reasoningOpenHands
69.85%
Not reportedClosed
48Grok 4.5xAIOpenHands · high reasoningOpenHands
69.00%
Not reportedClosed
49GPT-5.4OpenAIOpenHands · xhigh reasoningOpenHands
67.42%
Not reportedClosed
50GPT-5.5 FactoryOpenAIFactoryFactory
67.39%
Not reportedUnknown
51Beam 1.0ReflectionOpenHandsOpenHands
66.44%
Not reportedUnknown
52Qwen3.8-27BAlibabaOpenHands · xhigh reasoningOpenHands
64.85%
Not reportedOpen
53Qwen3.8 MaxAlibabaOpenHandsOpenHands
64.70%
Not reportedOpen
54Gemini 3.6 FlashGoogleOpenHands · high reasoningOpenHands
64.00%
Not reportedClosed
55GLM-5.2Z.AIOpenHands · max reasoningOpenHands
63.96%
Not reportedOpen
56GPT-5.3 CodexOpenAIOpenHands · xhigh reasoningOpenHands
61.77%
Not reportedClosed
57Mercury 3InceptionOpenHands · high reasoningOpenHands
60.93%
Not reportedClosed
58GPT-5.5 CodexOpenAICodexCodex
58.19%
Not reportedUnknown
59Claude Opus 4.6AnthropicOpenHandsOpenHands
57.57%
Not reportedClosed
60Claude Sonnet 4.6 Claude CodeAnthropicClaude CodeClaude Code
55.77%
Not reportedUnknown
61GPT-5.2OpenAIOpenHands · xhigh reasoningOpenHands
53.50%
Not reportedClosed
62Claude Opus 4.6 (Adaptive)AnthropicOpenHandsOpenHands
53.50%
Not reportedClosed
63Claude Sonnet 4.6AnthropicOpenHandsOpenHands
51.48%
Not reportedClosed
64DeepSeek V4 Pro 0813DeepSeekOpenHands · max reasoningOpenHands
49.93%
Not reportedOpen
65Composer 2.5CursorCursor CLICursor CLI
49.61%
Not reportedClosed
66Gemini 3.5 FlashGoogleOpenHands · high reasoningOpenHands
48.68%
Not reportedClosed
67GPT-5.4OpenAICodexCodex
48.47%
Not reportedClosed
68GPT-5.4 miniOpenAIOpenHands · xhigh reasoningOpenHands
47.97%
Not reportedClosed
69Qwen3.7 MaxAlibabaOpenHandsOpenHands
47.67%
Not reportedClosed
70MiniMax M3MiniMaxOpenHandsOpenHands
47.57%
Not reportedOpen
71Kimi K2.7 CodeMoonshot AIOpenHandsOpenHands
47.21%
Not reportedOpen
72Qwen3.7 PlusAlibabaOpenHandsOpenHands
46.39%
Not reportedClosed
73MiMo-V2.5XiaomiOpenHandsOpenHands
42.17%
Not reportedClosed
74GPT-5.2-CodexOpenAIOpenHands · high reasoningOpenHands
37.91%
Not reportedClosed
75Kimi K2.6Moonshot AIOpenHandsOpenHands
37.89%
Not reportedOpen
76Gemini 3.5 Flash-LiteGoogleOpenHands · high reasoningOpenHands
37.16%
Not reportedClosed
77MiMo-V2.5-ProXiaomiOpenHandsOpenHands
34.11%
Not reportedClosed
78Gemini 3.1 Pro PreviewGoogleOpenHands · high reasoningOpenHands
32.03%
Not reportedUnknown
79GLM-5.1Z.AIOpenHandsOpenHands
31.46%
Not reportedOpen
80GLM-5.3-FlashZ.AIOpenHands · max reasoningOpenHands
30.76%
Not reportedOpen
81GPT-5.4 nanoOpenAIOpenHands · high reasoningOpenHands
26.10%
Not reportedClosed
82Qwen3.6 PlusAlibabaOpenHandsOpenHands
25.57%
Not reportedClosed
83GPT-5.1OpenAIOpenHands · high reasoningOpenHands
24.61%
Not reportedClosed
84GLM 5 ThinkingZhipu AIOpenHandsOpenHands
23.36%
Not reportedUnknown
85Claude Sonnet 4.5 ThinkingAnthropicOpenHandsOpenHands
22.62%
Not reportedClosed
86GPT-5.1-Codex-MaxOpenAIOpenHands · high reasoningOpenHands
22.17%
Not reportedClosed
87Claude Opus 4.5 ThinkingAnthropicOpenHandsOpenHands
20.63%
Not reportedClosed
88Gemini 3 Flash PreviewGoogleOpenHands · high reasoningOpenHands
20.20%
Not reportedUnknown
89GPT-5OpenAIOpenHands · high reasoningOpenHands
20.09%
Not reportedUnknown
90Muse SparkMetaOpenHandsOpenHands
19.67%
Not reportedClosed
91Grok 4.3xAIOpenHands · high reasoningOpenHands
19.40%
Not reportedClosed
92InklingThinking Machines LabOpenHands · 0.99 reasoningOpenHands
19.21%
Not reportedOpen
93Inkling-SmallThinking Machines LabOpenHands · 0.99 reasoningOpenHands
19.05%
Not reportedOpen
94Kimi K2.5 ThinkingMoonshot AIOpenHandsOpenHands
17.54%
Not reportedUnknown
95Qwen3.5 Plus ThinkingAlibabaOpenHandsOpenHands
15.74%
Not reportedUnknown
96MiniMax M2.5MiniMaxOpenHandsOpenHands
14.85%
Not reportedClosed
97Gemini 3 Pro PreviewGoogleOpenHands · high reasoningOpenHands
14.30%
Not reportedUnknown
98GPT-5 miniOpenAIOpenHands · high reasoningOpenHands
14.17%
Not reportedClosed
99Grok Build 0.1xAIGrok BuildGrok Build
13.35%
Not reportedClosed
100GPT-5.1-CodexOpenAIOpenHands · high reasoningOpenHands
13.12%
Not reportedClosed
101Qwen3.6-27BAlibabaOpenHandsOpenHands
11.94%
Not reportedOpen
102MiniMax M2.7MiniMaxOpenHandsOpenHands
11.93%
Not reportedOpen
103Claude Haiku 4.5 ThinkingAnthropicOpenHandsOpenHands
11.39%
Not reportedClosed
104Laguna M.1PoolsideOpenHandsOpenHands
11.04%
Not reportedClosed
105Nemotron 3 Ultra 550b A55bNvidiaOpenHandsOpenHands
7.64%
Not reportedUnknown
106Swe 1.6 FastDevinDevin CLIDevin CLI
7.22%
Not reportedUnknown
107Laguna XS.2PoolsideOpenHandsOpenHands
5.21%
Not reportedOpen
108DeepSeek V3p2 ThinkingFireworks AIOpenHands · high reasoningOpenHands
5.11%
Not reportedUnknown
109Grok 4.20 0309 ReasoningxAIOpenHandsOpenHands
4.06%
Not reportedUnknown
110Mercury 2.5InceptionOpenHands · high reasoningOpenHands
3.68%
Not reportedClosed
111Qwen3 MaxAlibabaOpenHandsOpenHands
3.51%
Not reportedClosed
112Ling 3.0 Flash Af Rc3AntOpenHandsOpenHands
3.40%
Not reportedUnknown
113GLM-4.6Z.AIOpenHandsOpenHands
3.09%
Not reportedOpen
114Ling 3.0 Flash 2607AntOpenHandsOpenHands
2.91%
Not reportedUnknown
115Mistral Medium 3.5Mistral AIOpenHands · high reasoningOpenHands
2.89%
Not reportedUnknown
116Grok 4.1 Fast (Reasoning)xAIOpenHandsOpenHands
1.20%
Not reportedClosed
117Gemini 2.5 ProGoogleOpenHandsOpenHands
0.40%
Not reportedClosed
118Gemini 3.1 Flash Lite PreviewGoogleOpenHands · high reasoningOpenHands
0.00%
Not reportedUnknown
119Nemotron Lightning 3p5 30b A3bFireworks AIOpenHandsOpenHands
0.00%
Not reportedUnknown
120Grok 4 Fast (Reasoning)xAIOpenHandsOpenHands
0.00%
Not reportedClosed
121Mistral Small 2603Mistral AIOpenHands · high reasoningOpenHands
0.00%
Not reportedUnknown

How Vibe Code Bench v1.1 is shown here

BenchLM mirrors the public Vals AI Vibe Code Bench v1.1 leaderboard captured from https://www.vals.ai/benchmarks/vibe-code and updated by Vals on October 27, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.

Vibe Code Bench v1.1 is display only on BenchLM. Vals proprietary or Vals-hosted aggregate views are useful context, but BenchLM does not use them as weighted ranking inputs or as a replacement for benchmark-native source records.

Snapshot

121 Vals rows1 task viewsprivate datasetTasks: OverallDisplay only
Vibe Code Bench v1.1 on Vals AIVals methodologyVals home

The published Vibe Code Bench snapshot places Claude Sonnet 5.5 first at 92.39%. The third row is 1.95 points behind. The broader top-10 range is 3.99 points, so many of the published results sit in a relatively narrow band.

121 models have been evaluated on Vibe Code Bench. The benchmark falls in the Coding category. Vibe Code Bench is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About Vibe Code Bench

Year

2026

Tasks

End-to-end web application builds

Format

Full-stack app implementation benchmark

Difficulty

End-to-end software delivery

Vibe Code Bench v1.1 asks models to build full web apps with services such as Supabase, Stripe test mode, email, browsing, and file editing available. The score is overall application pass accuracy across private end-to-end app tasks.

Vibe Code Bench: Evaluating AI Models on End-to-End Web Application DevelopmentPublic benchmark source

Freshness and provenance

Version

Vibe Code Bench 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does Vibe Code Bench measure?

Vals.ai benchmark for evaluating whether models can build complete web applications from natural language specifications in a production-like development environment.

Which model leads the published Vibe Code Bench snapshot?

Claude Sonnet 5.5 currently leads the published Vibe Code Bench snapshot with 92.39% vibe code score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on Vibe Code Bench?

The October 27, 2026 snapshot contains 121 AI models.

Last updated: October 27, 2026 · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 6,100+ subscribers.

One email each week. Unsubscribe anytime.