Qwen Intelligence launches three mobile agents and opens its benchmarks
Source
github.com
Author
Qwen
Date
Why it matters
Qwen-Planner-AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → 27B reports 77.05% on MobilePA-Bench at an estimated $2.41 per 1,000 tasks in output tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition →, and the open MobileWorld and MobilePA-Bench suites give agent builders reproducible mobile-agent evalA repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.Full definition →. Scores are self-reported.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
postQwen3.8 Max goes live with 2.4T parameters, open weights promisedOpenRouter
postQwen Ships Omni-Modal Model With Native Audio-Video Agentic Tool UseAlibaba_Qwen
articleBenchmarking General Mobile Assistants in Challenging Real-World ScenariosYiqi Zhu, Feiyu Gao, Jiaxing Fan, Jiahui Zeng, Minggang Wu, Chenliang Li, Haiyang Xu, Peng Li, Ming Yan, Yang Liu