Vibe Code Bench 1-100 leaderboard
As of 6 Nov 2026, the top score on Vibe Code Bench 1-100 is 34.0% by Muse Wonder Full (Meta), ahead of Claude Opus 5.5 at 30.4%, across 26 model configurations with a published result (Vals AI, result dated 29 Oct 2026). Starting from a working app, the agent applies a hundred successive feature requests; the score is how many it completes without breaking the app. Vibe Code Bench 1-100 is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 6 Nov 2026)
- Muse Wonder Full holds the top Vibe Code Bench 1-100 score at 34.0% (Vals AI, 29 Oct 2026).
- Claude Opus 5.5 is second on Vibe Code Bench 1-100 at 30.4% (Vals AI, 29 Oct 2026).
- Vibe Code Bench 1-100: Extending a working web app, request by request, one hundred times. Run by Vals AI (Vals AI).
- 26 model configurations have a published Vibe Code Bench 1-100 result on BenchLeader (BenchLeader, 6 Nov 2026).
- Vibe Code Bench 1-100 does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Extending a working web app, request by request, one hundred times. Run by Vals AI.
- Published by
- Vals AI
- Category
- Coding
- Index weight
- Reference only
- Models
- 26
- Data as of
- 6 Nov 2026
Vals AI (vals.ai).
What the test looks like
Starting from a working app, the agent applies a hundred successive feature requests; the score is how many it completes without breaking the app.
How it is scored
Requests completed, run by Vals AI.
What to keep in mind
Tests sustained agentic coding rather than one-shot generation.
- 1Muse Wonder Full (max)34.0%
- 2Claude Opus 5.530.4%
- 3Claude Opus 528.5%
- 4Claude Fable 5.128.0%
- 5GPT-6 Astra (max)27.6%
- 6Muse Wonder (max)26.6%
- 7Qwen3.8 Max Internal24.8%
- 8GPT-5.6 Luna (max)22.6%
- 9Gemini 4 Argon (high)21.2%
- 10Muse Spark 1.3 (max)20.5%
- 11Claude Haiku 5.520.4%
- 12GPT-5.6 Sol (max)20.0%
- 13GLM 5.3 (max)20.0%
- 14Gemini 3.8 Flash (high)18.8%
- 15Kimi K3 (max)18.2%
| # | |||||
|---|---|---|---|---|---|
| 1 |  Muse Wonder Fullmax | 34.0% | 62.9 | $19.54 | 2026-10-29 |
| 2 |  Claude Opus 5.5 | 30.4% | 63.4 | $44.12 | 2026-10-29 |
| 3 | Claude Opus 5 | 28.5% | 65.2 | $41.49 | 2026-10-29 |
| 4 | Claude Fable 5.1 | 28.0% | 66.4 | $149.66 | 2026-10-29 |
| 5 |  GPT-6 Astramax | 27.6% | 67.8 | $84.72 | 2026-10-29 |
| 6 | Muse Wondermax | 26.6% | 62.3 | $13.72 | 2026-10-29 |
| 7 |  Qwen3.8 Max Internal | 24.8% | – | $11.59 | 2026-10-29 |
| 8 | GPT-5.6 Lunamax | 22.6% | 56.3 | $4.91 | 2026-10-29 |
| 9 |  Gemini 4 Argonhigh | 21.2% | 66.0 | $18.66 | 2026-10-29 |
| 10 | Muse Spark 1.3max | 20.5% | 62.0 | $5.60 | 2026-10-29 |
| 11 | Claude Haiku 5.5 | 20.4% | – | $8.30 | 2026-10-29 |
| 12 | GPT-5.6 Solmax | 20.0% | 65.4 | $47.43 | 2026-10-29 |
| 13 |  GLM 5.3maxopen ↗ | 20.0% | 59.0 | $8.24 | 2026-10-29 |
| 14 | Gemini 3.8 Flashhigh | 18.8% | 62.0 | $18.91 | 2026-10-29 |
| 15 |  Kimi K3maxopen ↗ | 18.2% | 62.1 | $9.40 | 2026-10-29 |
| 16 |  DeepSeek V4 Promaxopen ↗ | 17.5% | 57.0 | $5.89 | 2026-10-29 |
| 17 | DeepSeek V4.1 Flashhighopen ↗ | 16.4% | 55.7 | $0.688 | 2026-10-29 |
| 18 | GLM 5.3 Flashmaxopen ↗ | 16.0% | 54.2 | $0.926 | 2026-10-29 |
| 19 | GPT-5.6 Terramax | 14.8% | 60.7 | $33.33 | 2026-10-29 |
| 20 |  Grok 4.6high | 14.8% | 60.5 | $26.27 | 2026-10-29 |
| 21 |  Mistral Large 4highopen ↗ | 14.3% | 49.5 | $3.44 | 2026-10-29 |
| 22 | Claude Sonnet 5 | 13.8% | 55.9 | $71.15 | 2026-10-29 |
| 23 | Qwen3.8 Maxmax | 12.8% | 58.8 | $9.68 | 2026-10-29 |
| 24 |  MiniMax M3open ↗ | 9.2% | 56.4 | $3.03 | 2026-10-29 |
| 25 | Inklingopen ↗ | 7.3% | 47.5 | $5.00 | 2026-10-29 |
| 26 | Gemini 3.1 Prohigh | 6.7% | 60.1 | $3.36 | 2026-10-29 |
More coding benchmarks · All LLM benchmarks
Related benchmarks
Top head-to-heads on Vibe Code Bench 1-100
Cite as: BenchLeader, “Vibe Code Bench 1-100 leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/vals_vibe_code_bench.
Vibe Code Bench 1-100: questions
- What does Vibe Code Bench 1-100 measure?
- Starting from a working app, the agent applies a hundred successive feature requests; the score is how many it completes without breaking the app. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on Vibe Code Bench 1-100?
- As of 6 Nov 2026, the highest score on Vibe Code Bench 1-100 is 34.0% by Muse Wonder Full (result dated 29 Oct 2026), ahead of Claude Opus 5.5 at 30.4%. Scores come from Vals AI and are re-read daily.
- Which AI model leads Vibe Code Bench 1-100?
- Muse Wonder Full leads Vibe Code Bench 1-100 with 34.0% as of 6 Nov 2026 (result dated 29 Oct 2026), ahead of Claude Opus 5.5 at 30.4%.
- How many models have Vibe Code Bench 1-100 results?
- 26 model configurations have a Vibe Code Bench 1-100 result on BenchLeader, all taken from Vals AI.
- Who runs Vibe Code Bench 1-100 and how often is it updated?
- Vibe Code Bench 1-100 is published by Vals AI. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does Vibe Code Bench 1-100 count toward the BenchLeader Index?
- No. Vibe Code Bench 1-100 is shown for reference but left out of the composite index.