Kaggle BackTrace Backend Debugging: leaderboard
Community benchmark on Kaggle by himanshunakrani: 16 backend incident scenarios described through logs and metrics, such as missing indexes, retry loops, cache stampedes, Kafka partition skew and connection-pool leaks. A scenario passes when a judge model accepts every root-cause and remedy criterion; percent of scenarios passed, higher is better.
Metric: Scenarios Passed (%). Source: www.kaggle.com. 36 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 Mini | 81.25 | #228 |
| 2 | Gemini 2.5 Pro | 68.75 | #159 |
| 3 | Claude Sonnet 4.5 | 68.75 | #148 |
| 4 | Claude Sonnet 4.6 | 68.75 | #94 |
| 5 | GPT-5.4 | 68.75 | #83 |
| 6 | Claude Opus 4.6 | 68.75 | #67 |
| 7 | Claude Haiku 4.5 | 68.75 | #287 |
| 8 | GPT-OSS-20B | 68.75 | #523 |
| 9 | DeepSeek V3.1 | 68.75 | #275 |
| 10 | Qwen 3 235B A22B 2507 Instruct | 68.75 | #315 |
| 11 | Gemma 4 26B A4B | 68.75 | #360 |
| 12 | Gemini 2.5 Flash | 62.5 | #255 |
| 13 | GPT-5.5 | 62.5 | #33 |
| 14 | Claude Opus 4.8 | 62.5 | #39 |
| 15 | Gemini 3 Flash (Preview) | 62.5 | #90 |
Interactive version: theaggregate.ai/benchmark?slug=kaggle-backtrace-backend-debugging · How It Works · Data refreshed daily, snapshot 2026-11-05.