Inference
EN
Vals AI

Vibe Code Bench 1-100

Can models extend a working web application across a long sequence of dependent requests?

ModelsModels with a result.26
Top resultBest result on this benchmark.34.0%Muse Wonder Full
Top-3 spreadPoints from first to third.5.4 pts
UpdatedDate the source changed the results.29 Oct 2026

Result and cost

26 of 26 results

Results

26 results
#ModelResultThe score from the source.CostUS dollars to run one task.TimeTime to run one task.
01Muse Wonder FullOpenHands34.0% ±5.3$19.541.9 h
02Claude Opus 5.5OpenHands30.4% ±4.8$44.122.5 h
03Claude Opus 5OpenHands28.5% ±4.3$41.4982 min
04Claude Fable 5.1OpenHands28.0% ±4.5$149.662.9 h
05GPT-6 AstraOpenHands27.6% ±4.1$84.723.4 h
06Muse WonderOpenHands26.6% ±4.3$13.7231 min
07Qwen3.8 Max InternalOpenHands24.8% ±4.5$11.592.6 h
08GPT-5.6 LunaOpenHands22.6% ±4.0$4.912.0 h
09Gemini 4 ArgonOpenHands21.2% ±3.8$18.661.7 h
10Muse Spark 1.3 MaxOpenHands20.5% ±3.8$5.6020 min
11Claude Haiku 5.5OpenHands20.4% ±3.8$8.304.1 h
12GPT-5.6 SolOpenHands20.0% ±3.6$47.432.6 h
12GLM-5.3OpenHands20.0% ±3.8$8.2486 min
14Gemini 3.8 FlashOpenHands18.8% ±3.3$18.9166 min
15Kimi K3OpenHands18.2% ±3.9$9.401.8 h
16DeepSeek V4 Pro 0813OpenHands17.5% ±3.4$5.8938 min
17DeepSeek V4.1 Flash 1017OpenHands16.4% ±3.2$0.68828 min
18GLM-5.3-FlashOpenHands16.0% ±3.3$0.92639 min
19GPT-5.6 TerraOpenHands14.8% ±3.1$33.331.8 h
19Grok 4.6OpenHands14.8% ±3.2$26.272.2 h
21Mistral Large 4OpenHands14.3% ±3.2$3.4460 min
22Claude Sonnet 5OpenHands13.8% ±3.0$71.154.3 h
23Qwen3.8 MaxOpenHands12.8% ±3.2$9.681.6 h
24MiniMax M3OpenHands9.2% ±2.9$3.0326 min
25InklingOpenHands7.3% ±1.8$5.0050 min
26Gemini 3.1 Pro PreviewOpenHands6.7% ±2.4$3.3611 min