RadixArk releases DaisyRose-27B, an RL post-trained Qwen3.8-27B agent model
DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency
RadixArk released DaisyRose-27B, its first model, built on Qwen3.8-27B and trained with reinforcement learning using Miles on 1,200 curated long-horizon agent tasks in 18 days on 144 H200 GPUs. Terminal-Bench 4.1 pass@1 rose from 5.6% to 12.0%, with smaller gains on Terminal Bench 2.1, DeepSWE v1.1, SWE Bench Pro v2, HLE, IFBench and GPQA Diamond. Checkpoints are available under Apache 2.0 in BF16, FP8 and NVFP4, and the model keeps Qwen3.8-27B's architecture, tokenizer, image understanding and 256K-token context window.
RadixArk's first model shows how a small curated task set and an asynchronous RL backend can lift long-horizon agent performance on a 27B base model.
Source:Hugging Face Blog · huggingface.co