Coding benchmark
ProgramBench leaderboard
ProgramBench: Can Language Models Rebuild Programs From Scratch?. Every model the catalog carries a published ProgramBench value for, ranked by that value.
A cleanroom software-engineering benchmark where agents receive only a compiled executable and documentation, then must architect and implement a complete codebase that reproduces the original program's behavior.
ProgramBench ranking
8 models with a published ProgramBench value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Cleanroom executable reimplementation |
|---|---|---|---|
| 1 |  Claude Opus 5 | Anthropic | 93.0 |
| 2 | Claude Fable 5.1 | Anthropic | 87.6 |
| 3 |  Kimi K3 | Moonshot AI | 77.8 |
| 4 |  GLM-5.2 | Z.AI | 63.7 |
| 5 | Kimi K2.7 Code | Moonshot AI | 53.6 |
| 6 |  DeepSeek V4.1 Flash | DeepSeek | 20.3 |
| 7 | GLM-5.3 | Z.AI | 19.0 |
| 8 |  Hy4 preview | Tencent | 17.5 |
Evidence key: Observed
Rows are ordered by the value ProgramBench: Can Language Models Rebuild Programs From Scratch? published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.