CodeEditorBench Plus - Code Debug (Zero-shot): leaderboard
Metric: pass@1 (%, greedy decoding, zero-shot, CodeEditorBench_Plus). Source: codeeditorbench.github.io. Saturation forecast: Estimated already saturated. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4 (0613) | 31.6 |
| 2 | GPT-3.5 Turbo (1106) | 29 |
| 3 | Gemini 1.0 Pro | 28.6 |
| 4 | Phind-CodeLlama-34B-v2 | 23 |
| 5 | GLM-4 | 22 |
| 6 | CodeLlama-13B-Instruct-hf | 17.6 |
| 7 | Magicoder-S-CL-7B | 17.4 |
| 8 | CodeLlama-34B-hf | 16.3 |
| 9 | WizardCoder-15B-V1.0 | 15.9 |
| 10 | CodeLlama-7B-Instruct-hf | 15.5 |
| 11 | CodeLlama-34B-Instruct-hf | 13.1 |
Interactive version: theaggregate.ai/benchmark?slug=codeeditorbench-plus-code-debug-zero-shot · How It Works · Data refreshed daily, snapshot 2026-09-26.