ArchBench Serverless Component Generation - CodeBLEU: leaderboard
CodeBLEU similarity (0-100) between a model's generated serverless function and the original implementation, without human edits and with the paper's most detailed prompt; higher is better (ArchBench, IIIT Hyderabad, 2025).
Metric: CodeBLEU (0-100). Source: github.com. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 | 35 | #996 |
| 2 | GPT-3.5 Turbo | 32 | #2773 |
| 3 | CodeQwen1.5-7B Chat | 30 | #5781 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=archbench-serverless-component-generation-codebleu · How It Works · Data refreshed daily, snapshot 2026-11-12.