QwenPaw Releases

Major provider support release adding OpenRouter and OpenCode providers, Matrix channel rewrite, and enhanced security guards.

Apr 14, 2026

This release focuses on stability and bug fixes across various UI components, including Dialog, Drawer, and Popover. It addresses hydration issues, hot module replacement (HMR) errors, and improves accessibility standards.

Apr 13, 2026

This release marks a significant milestone with the official rebranding of CoPaw to QwenPaw, reflecting a deeper integration with the Qwen open-source ecosystem and the personal assistant mission.

Apr 12, 2026

A comprehensive feature release introducing plugins support, new model providers (SiliconFlow, Zhipu), enhanced security, and significant UI improvements including chat search, pinned sessions, and skill commands.

Apr 9, 2026

Feature-rich release adding multimodal video support, new channels (OneBot v11, DingTalk AI Cards), UI modernization with dark mode system preference, and extensive bug fixes across channels and tools.

Apr 3, 2026

Post-release patch with style improvements, Feishu DONE reaction support, DingTalk AI Card enhancements, and various bug fixes including WeCom WebSocket reliability and skill management improvements.

Apr 1, 2026

Post-release patch focusing on UI improvements, MCP console refresh, performance optimizations for skill lists, and various bug fixes including WeCom reconnection and provider improvements.

Mar 31, 2026

Initial post-release patch addressing various bugs including console UI issues, skill requirement parsing, provider fixes, and documentation improvements following the major v1.0.0 release.

Mar 31, 2026

Major 1.0.0 release introducing multi-agent system support, CoPaw Local model provider, enhanced security features, extensive UI improvements, and comprehensive documentation updates.

Mar 30, 2026

This release focuses on refactoring the tokenizer to use skeletoken, removing legacy dependencies, deprecating Python 3.9, and updating benchmark results to MTEB V2.

Mar 26, 2026

Post-release patch for v0.2.0 introducing console chats, agent enable/disable toggle, Feishu widescreen card support, browser multi-workspace cookie management, and various bug fixes.

Mar 25, 2026

Feature release introducing inter-agent communication, built-in QA agent, enhanced security features, Feishu SDK migration, audio/video/speech support, and extensive UI improvements.

Mar 24, 2026

Initial release with various bug fixes and improvements including Anthropic retry handling, token usage fixes, Swagger documentation routing, and various platform-specific fixes.

Mar 20, 2026

The initial release introducing the core Multi-Agent/Multi-Workspace architecture, robust security features including a skill scanner, multimodal capabilities (images/audio), and support for new LLM providers like Gemini and DeepSeek.

Mar 19, 2026

Release notes are hosted externally on NVIDIA documentation. Specific details regarding changes are not included in the provided text.

Mar 18, 2026

This release introduces a new proxyproto plugin for Proxy Protocol support, enhances DNS logging metadata, and updates the Go version to 1.26.1 to address multiple security vulnerabilities.

Mar 6, 2026

This release directs users to external NVIDIA documentation for comprehensive release notes, as the changelog is hosted elsewhere.

Jan 16, 2026

This update focuses on bug fixes, specifically addressing ESM import compatibility and Vimeo caption visibility issues.

Jan 3, 2026

This release marks a significant accessibility improvement by open-sourcing the Voice-Pro codebase and removing all associated costs.

Dec 5, 2025

A model2vec release adding weight support to loss functions, fixing padding token recognition, adding Python 3.13 support, classifier freezing, vocabulary quantization, and configurable parameters.

Oct 5, 2025