"KAT-Coder-V2.5 Technical Report"
This new model from KwaiKAT called KAT-Coder-V2.5 rivals GLM-5.2 on coding, and in their report, they argue that better coding agents need better training worlds, not just bigger models.
To train their model, they built this AutoBuilder which turns real repos into reproducible sandboxes with real tests, then converts messy issues into clear executable tasks.
It recovers useful failed attempts, filters bad passing solutions, and randomizes tool harnesses so the model learns software engineering, not one interface.
The results in top PinchBench performance and second place behind Opus 4.8 on repo-level SWE.
Teach a man to code, and he'll be unemployed for the rest of his life.
Teach a man to prompt, and he'll be unemployed for the rest of his life AND have a crushing debt for token usage.