Vibe Code Bench v1.1
Code · 2026-03-02
Web-app generation benchmark testing whether models can build usable applications from natural-language prompts.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Sonnet 5.5 | 92.4 |
| Claude Haiku 5.5 | 90.4 |
| Claude Fable 5 | 90.4 |
| Gemini 4 Carbon | 89.1 |
| GPT-6.1 Sol | 88.9 |
| Claude Opus 5 | 88.4 |
| GPT-6 Sol | 87.8 |
| DeepSeek V4.1 Flash 1017 | 87.7 |
| Grok 4.7 | 86.2 |
| Muse Wonder Full | 86.0 |
| Muse Spark 1.3 | 85.9 |
| MiMo-V2.6-Pro | 85.2 |
| Kimi K3 | 85.0 |
| Kimi K3.1 Pro | 84.9 |
| DeepSeek V4.1 Flash | 84.7 |
| GPT-6.1 Terra | 84.0 |
| Ember-1 | 83.9 |
| Muse Wonder | 83.1 |
| Opus 4.8 | 82.7 |
| DeepSeek V4 Pro 0813 | 82.3 |
| GPT-6 Luna | 81.6 |
| Sonnet 5 | 81.3 |
| GPT-5.6 Sol | 80.5 |
| Muse Spark 1.2 | 79.1 |
| MiMo-V2.6-Flash | 79.0 |
| Gemini 3.8 Flash | 78.7 |
| Mistral Large 4 Preview | 78.4 |
| GLM-5.3 | 78.1 |
| Hy4 Preview | 77.5 |
| Qwen3.8 Max (0902) | 77.1 |
| GPT-5.6 Luna | 77.1 |
| Grok 4.6 | 76.2 |
| Muse Spark Open | 76.2 |
| DeepSeek-V4-Flash-0731 | 74.7 |
| GPT-5.6 Terra | 74.6 |
| MiniMax M3.1 Flash | 73.9 |
| Muse Spark 1.1 | 72.2 |
| Opus 4.7 | 71.0 |
| Step 5 Preview | 70.4 |
| Gemini 3.7 Flash | 70.4 |