Definition
Plain language
A strong screen-reading phone agent used as a comparison point.
As stated in the literature
A GUI-based mobile agent reporting around 77.6% on AndroidWorld; set aside in the terminal-agent comparison because its evaluation pipeline isn't publicly reproducible.
Why it matters: It serves as a strong screen-reading reference point, though its results can't be independently reproduced without a public evaluation pipeline.
For example, it reads a phone's screen and taps through menus to complete a task, scoring around 77.6% on one benchmark.
Heard on the show
“There's a system called UI-Venus that hits seventy-seven-point-six percent on AndroidWorld — higher than the terminal agent's seventy-one-point-eight.”Episode 157 — When an AI Coding Agent Drives a Phone Through the Terminal, No Screen Needed