Bug Hunt Bench: leaderboard
Coding models find and repair 105 planted defects in two production repositories: 45 in a VS Code extension and 60 in a learning-management system. Blind grading counts verified planted fixes. Reasoning effort, the July DeepSeek revision and the retired Gemini CLI protocol remain distinguishable; superseded runs follow the publisher’s replacement markers. This board reports Bug Hunt Bench using Planted Bugs Fixed (out of 105).
Metric: Planted Bugs Fixed (out of 105). Source: bughunt.productcompass.pm. Saturation forecast: Around September 2028. 132 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 5.5 (Max) | 51.3 | #14 (Claude Sonnet 5.5) |
| 2 | GPT-6 Astra (Max) | 45 | #6 (GPT-6 Astra) |
| 3 | GPT-6.1 Sol (Max) | 44.3 | #7 (GPT-6.1 Sol) |
| 4 | GPT-5.6 Sol (Max) | 43.5 | #22 (GPT-5.6 Sol) |
| 5 | Claude Fable 5.1 (Max) | 43 | #10 (Claude Fable 5.1) |
| 6 | GPT-6 Astra (xHigh) | 43 | #6 (GPT-6 Astra) |
| 7 | DeepSeek V4.1 Flash (1017) (Max) | 42.7 | #27 (DeepSeek V4.1 Flash (1017)) |
| 8 | GPT-6.1 Sol (xHigh) | 42.7 | #7 (GPT-6.1 Sol) |
| 9 | Muse Wonder Full (Max) | 42 | #12 (Muse Wonder Full) |
| 10 | Claude Opus 5.5 (Max) | 41.7 | #3 (Claude Opus 5.5) |
| 11 | Gemini 4 Argon (High) | 40.7 | #9 (Gemini 4 Argon) |
| 12 | GPT-6.1 Terra (Max) | 40.7 | #15 (GPT-6.1 Terra) |
| 13 | DeepSeek V4.1 Flash (1017) (High) | 40.4 | #27 (DeepSeek V4.1 Flash (1017)) |
| 14 | Gemini 4 Argon (Medium) | 39.3 | #9 (Gemini 4 Argon) |
| 15 | GPT-5.6 Sol (xHigh) | 39 | #22 (GPT-5.6 Sol) |
Interactive version: theaggregate.ai/benchmark?slug=bug-hunt-bench · How It Works · Data refreshed daily, snapshot 2026-11-05.