CodeSpecBench-Repo - Completeness: leaderboard
CodeSpecBench (2026) asks models to write executable precondition and postcondition functions that specify the intended behaviour of a program; the repository-level set uses the 500 SWE-bench Verified issues, injecting the specifications around the issue-relevant functions. Completeness, the share of specifications that reject every invalid test, in percent; higher is better.
Metric: Completeness (%), the share of generated executable pre/postcondition specifications that reject every invalid test, on CodeSpecBench-Repo (500 SWE-bench Verified issues; specifications injected around the issue-relevant functions, trigger tests on the fixed and buggy versions with UTBoost-augmented tests); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 15 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5 | 58.6 | #178 | |
| 2 | Claude Sonnet 4.5 | Thinking | 57.2 | #261 |
| 3 | Gemini 2.5 Pro | 53 | #276 | |
| 4 | GPT-5 Mini | 45.8 | #364 | |
| 5 | GPT-OSS-120B | 36.2 | #704 | |
| 6 | GPT-OSS-20B | 31.6 | #1183 | |
| 7 | DeepSeek V3.2 | Non-reasoning | 30.2 | #419 |
| 8 | QwQ-32B | 25 | #907 | |
| 9 | Qwen 3 14B | 17.8 | #1269 | |
| 10 | Gemini 2.5 Flash | 17.6 | #463 | |
| 11 | Qwen 3 32B | 15.6 | #972 | |
| 12 | Qwen 3 8B | Non-reasoning | 10 | #1873 |
| 13 | Qwen 3 4B | 1.2 | #2348 | |
| 14 | Qwen 3 1.7B | Non-reasoning | 0.2 | #5860 |
| 15 | Qwen 3 0.6B | 0 | #8350 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=codespecbench-repo-completeness · How It Works · Data refreshed daily, snapshot 2026-11-21.