RadixArk Ships DaisyRose-27B, a 27B Agentic Coding Model Trained With Reinforcement Learning Only
RadixArk has released DaisyRose-27B, an open-source coding model built upon Qwen3.8-27B that was refined exclusively through reinforcement learning. This model demonstrates improved performance on the Terminal Bench 4.1 coding benchmark while maintaining a small enough footprint to operate on a single GPU.
ModelsDaisyRose 27B
Covered by 1 source
- MMarkTechPost↗Asif Razzaq1d ago