A native macOS AI chat app powered by MLX. 100% local inference on Apple Silicon, no cloud required. Built with ShipSwift.
Stars
1.7k
Forks
138
Watchers
1.7k
Open Issues
20
Overall repository health assessment
No package.json found
This might not be a Node.js project
33
commits
Remove mlx-swift-lm performance analysis documents: Fork plan and performance research report deleted as part of project restructuring.
64bce8dView on GitHubfeat: Implement native tool calling with web search and file operations in Klee
5fae350View on GitHubfeat: Enhance chat functionality with action execution and real-time thinking updates
141b2cfView on GitHubfeat: Update model cache paths to use ~/.klee directory and add Qwen 3 8B model to recommended list
ac90d16View on GitHubAdd performance gap analysis and optimization plan for mlx-swift-lm
946ddeaView on GitHubfeat: Add performance metrics to LLMService and display in ChatView and ChatConfigView
aacb8aaView on GitHubRefactor Klee application structure: remove MCPServerListView, update SettingsView, and introduce ModuleManager and ChatConfigView
652bf71View on GitHubfeat: Remove MCP and Node.js dependencies, streamline architecture for a pure Swift implementation
689e96fView on GitHubRefactor code structure for improved readability and maintainability
21592fbView on GitHubfeat: Add InputBarView, MessageBubbleView, and WelcomeView for chat interface
e7915fcView on GitHubfeat: Enhance Klee's CLI integration and architecture proposal
153f613View on GitHubfeat: Implement tokenizer_config.json patching for missing chat_template in models
d8e9722View on GitHubdocs: Add acknowledgements section to README for ShipSwift contributions
00fb4f2View on GitHubRemove knowledgebase, marketplace, note routes, and related types; delete TypeScript configuration file
929b42fView on GitHubfeat: Update model configurations and enhance image support in chat
f405b59View on GitHub