A practical, safety-first guide to evaluating Android GUI agents with reproducible setup steps and model options.
No 24GB GPU? Use a deployed model service endpoint if you have access.
Trust anchors from official sources
Use official sources for code, models, and licensing. This site curates links and explains how to verify them.
Review the repository and run your own builds for reproducibility.
Download from Hugging Face or ModelScope (no re-hosting).
Apache-2.0 terms for the code, with models governed by their own licenses.
Not affiliated with the original authors; focused on safety and evaluation.
Option A is recommended for most users. Option B is for local power users.
No local GPU required. Use a deployed model service endpoint and pass --base-url.
Requires an NVIDIA GPU (24GB+ VRAM recommended), vLLM or SGLang, and ~20GB model files.
An independent, educational resource for testing Android GUI agents safely.
Step-by-step guidance for evaluating Open-AutoGLM without guesswork.
Open-source, reproducible flows that surface hardware and safety tradeoffs.
Not affiliated with the original authors; focused on objective evaluation.
Join the waitlist for guided Android regression testing workflows and agent-safe checklists.
Sponsor longform guides, deployment notes, or community spaces. Ads and Creem checkout are placeholders for now.