QwenLM Launches Open-Source Plugin System Bringing Multimodal AI Capabilities to Any Agent
Summary
QwenLM launches Qwen-MM-Plugins, an open-source plugin system that gives any AI agent powerful multimodal capabilities — including image, video, audio, and document understanding — with modular tools supporting 3D modeling, web search, video editing, and more, compatible with Claude Code, Codex, and Gemini CLI.
Key Points
- QwenLM releases Qwen-MM-Plugins, an open-source GitHub repository designed to give any AI agent native multimodal capabilities, including image, video, audio, and document understanding.
- The plugin system offers modular capabilities installed separately, covering local I/O, cloud APIs for vision and audio, web search, long-video memory, video editing, 3D modeling via Blender and FreeCAD, and educational video generation.
- A guided installer script supports multiple AI coding harnesses including Claude Code, Codex, and Gemini CLI, with configuration managed through a shared config file and dependencies handled automatically via uvx on first launch.