Description
Having AI order food or reply to messages on your phone fails when scripts break on UI changes. PhoneDriver uses the Qwen3-VL vision-language model to read the screen and complete natural-language tasks on Android via ADB.
It has a built-in Gradio web UI and runs open models locally.
Vision:Understands any UI.
Natural language:Say what you want.
ADB:Taps, swipes and typing.
Web UI:Gradio.
It has a built-in Gradio web UI and runs open models locally.
Features
Vision:Understands any UI.
Natural language:Say what you want.
ADB:Taps, swipes and typing.
Web UI:Gradio.
