extreme1228/ScaleCUA-qwen3-vl-scienceboard

ScaleCUA-qwen3-vl-scienceboard

A Qwen3-VL-8B computer-use agent from ScaleCUA, trained for ScienceBoard. Stage: online-RL (RLVR).

• —📄 Paper: ScaleCUA: Scaling Computer Use Agents with Verifiable Task Synthesis and Efficient Online RL
• —💻 Code: https://github.com/THUDM/SCALE-CUA
• —🤗 Collection: https://huggingface.co/collections/extreme1228/scalecua-6a4cad5e3c80a230622eaff3 · 🤖 ModelScope

What is this

This is the online-RL checkpoint (the released agent). Its SFT start point is `extreme1228/ScaleCUA-qwen3-vl-scienceboard-sft`. It reaches 51.9% on ScienceBoard.

ScaleCUA scales online RL for computer-use agents by pairing VeriGen (verifiable GUI-task synthesis) with Frontier Sampling and Visual Context Segmentation. See the repo for details.

Usage

Serve with an OpenAI-compatible server (e.g. vLLM) and drive it with the evaluation runtime in `osworld_eval`:

pip install -U vllm
python -m vllm.entrypoints.openai.api_server --model extreme1228/ScaleCUA-qwen3-vl-scienceboard \
  --served-model-name scalecua --trust-remote-code --port 8000

Then follow `REPRODUCE.md` to run ScienceBoard evaluation.

Results

51.9% on ScienceBoard (see the paper for the full table).

License

Weights are released for research; the model inherits the license of its base model (Qwen3-VL-8B). Repository code is under the licenses in the GitHub repo.

Citation

@misc{lv2026scalecua,
  title  = {ScaleCUA: Scaling Computer Use Agents with Verifiable Task Synthesis and Efficient Online RL},
  author = {Bowen Lv and Xiao Liu and Yanyu Ren and Hanyu Lai and Bohao Jing and Hanchen Zhang and Yanxiao Zhao and Shuntian Yao and Jie Tang and Yuxiao Dong},
  year   = {2026},
  url    = {https://github.com/THUDM/SCALE-CUA}
}