LM Studio 0.4.0
10 days ago
- #AI Deployment
- #LM Studio
- #REST API
- LM Studio 0.4.0 introduces parallel requests with continuous batching for high throughput serving.
- New non-GUI deployment option with llmster, allowing server-native operation without reliance on the GUI.
- Stateful REST API with /v1/chat endpoint for chatting with local models, supporting multi-step workflows.
- Refreshed user interface with features like Split View for multiple chat sessions and Developer Mode for advanced options.
- New CLI experience centered around the lms chat command for interactive chat sessions in the terminal.
- Support for concurrent inference requests to the same model with llama.cpp engine version 2.0.0.
- Export chats to PDF, markdown, or plain text.
- Permission keys introduced for controlling client access to the LM Studio server.
- Various bug fixes and improvements across builds, including UI updates, model search enhancements, and performance optimizations.