Domains: Coding
Skills: AgenticCodingTool UseLong ContextInstruction Following

Four-Factor Integrity Breakdown & Primary-Source Evidence

Privacy

Weight 40%
10/10
Access Model: Strictly Private
"The dataset contains 50 test set scenarios and 50 separate validation scenarios. All model scores reported on this page are based only on the test subset."
"The test and validation sets are separate. Validation results and historical v1 scores are not mixed into this leaderboard."

Originality

Weight 25%
10/10
Reuse Profile: All Original (0% reuse)
"The v1.1 workflow definitions contain 6022 workflows across 456 requests in 50 test scenarios. Of these, 1572 (26.1%) check earlier behavior, and 4450 cover the new request. These figures exclude the separate validation set."

Freshness

Weight 20%
10/10
Initial Release: 2026-09Latest Refresh: 2026-11Rolling Updates: No
"In v1.1, we changed iteration prompts to make requirements less ambiguous. We also audited task workflows to check that each assertion follows from the task request. We added a step to the grading process in which a second judge model, Claude Sonnet 5.5, reviews browser-agent failure verdicts against the recorded evidence."

Grading

Weight 15%
7.3/10
Integrity: Independent StandardizedSandbox: live internet
"All models use the same modified OpenHands V1 environment, with terminal access and preconfigured Supabase, Stripe test services, and MailHog. Each task has one ten-hour generation budget shared across its iterations. Evaluation time is excluded."

Frontier Model Standings on Vibe Code Bench 1-100 v1.1

Top 10 active models tracked
Rank Model Official Score Percentile
#1GPT-6.2 Astra max OpenAI55.2%100.0p
#2Gemini 4 Carbon high Google49.7%95.6p
#3GPT-6.2 Sol max OpenAI40.9%86.4p
#4Claude Opus 5.5 max Anthropic30.9%77.2p
#5GLM-5.3 max Z AI29.8%68.0p
#6MiMo-V2.6-Pro default Xiaomi26.4%58.8p
#7Grok 4.7 xhigh SpaceXAI26.4%49.6p
#8DeepSeek V4.1 Flash 1017 max DeepSeek26.1%40.4p
#9MiMo-V2.6-Flash default Xiaomi24.5%31.2p
#10MiniMax-M3.1-Flash max MiniMax19.8%22.0p

Embeddable Audit Badge

View Markdown (.md)

Benchmark maintainers can embed this live SVG badge in GitHub READMEs or documentation:

BenchAudit Vibe Code Bench 1-100 v1.1 Tier S 9.6/10
[![BenchAudit Tier S](https://benchaudit.org/badges/vcb-1-100-v1-1.svg)](https://benchaudit.org/benchmarks/vcb-1-100-v1-1/)