AI BriefingQwen Intelligence Launches Three Mobile AI Agents and Open Benchmark Suite
Key point
Qwen Intelligence launched three SOTA mobile AI agents, including a Mobile-Use Agent with a 90% end-to-end success rate and a Creative Agent generating images in 3 seconds, alongside an open benchmark suite.
Details
Qwen Intelligence has launched a suite of three state-of-the-art mobile AI agents designed to bring personal intelligence to everyday users. The platform introduces specialized agents for planning, execution, and creative generation, alongside an open benchmark suite to evaluate mobile AI performance.
Agent Capabilities and Performance
The launch features three distinct agents, each optimized for specific mobile tasks:
- Mobile Planner Agent: Focuses on planning, decomposing, and orchestrating complex tasks. It ranks #1 on MobilePA-Bench, MobilePA-Bench Business, and MobilePA-Bench Memory.
- Mobile-Use Agent: Designed to execute tasks using an API-first approach with GUI fallback. It achieves scores of 82.1 on MobileWorld, 92.2 on MobileWorld-Real, and 97.2 on AndroidDaily, with a 90% end-to-end success rate.
- Mobile Creative Agent: Converts single sentences into ready-to-use creations. It generates images in 3 seconds, approximately 2x faster than leading peers.
Open Benchmark Suite
Qwen is also releasing an open benchmark suite to standardize evaluation across planning, cross-app execution, real-device performance, and safety. The suite includes:
- MobilePA-Bench
- MobileWorld
- MobileWorld-Real
- MobileWorld-Safety
Resources for the agents and benchmarks are available via the Qwen Intelligence official website and associated GitHub repositories.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Apple Releases LensVLM-9B for Processing Compressed Text Images