GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
extreme1228
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
extreme1228
Model Tree
Input Modalities
Output Modalities
Supported Functionality
This is the SFT start-point checkpoint used to initialize online RL; the corresponding RL checkpoint is the released agent (see the collection).
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.
Serve with an OpenAI-compatible server (e.g. vLLM) and drive it with the
evaluation runtime in osworld_eval:
bash
pip install -U vllmpython -m vllm.entrypoints.openai.api_server --model extreme1228/ScaleCUA-qwen3.5-osworld-sft \--served-model-name scalecua --trust-remote-code --port 8000
Then follow REPRODUCE.md to run OSWorld evaluation.
(see the paper for the full table).
Weights are released for research; the model inherits the license of its base model (Qwen3.5-9B). Repository code is under the licenses in the GitHub repo.
bibtex
@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}}