LM Studio 0.4.0 introduces significant upgrades, including the llmster daemon, which separates core functionality from the GUI. This allows for headless deployments across various environments, such as cloud servers or local machines. Users can now run the daemon independently, making it suitable for terminal-based workflows. Key commands include starting the daemon, downloading models, and initiating chat sessions directly from the terminal. This flexibility enhances deployment options and caters to users who prefer non-GUI setups. The release also brings parallel requests with continuous batching, allowing multiple requests to be processed simultaneously rather than queued. Users can configure the maximum concurrent predictions in the model loader dialog. This feature is particularly useful for high-throughput scenarios, as it optimizes resource usage without significant overhead. The revamped UI enhances user experience with features like chat export options and a split view for side-by-side chat sessions. A new stateful REST API endpoint, `/v1/chat`, streamlines interactions with local models, enabling users to maintain conversation context through response IDs. This design keeps requests smaller and facilitates multi-step workflows. Alongside these improvements, the article highlights the introduction of permission keys for controlling client access to the LM Studio server, enhancing security and management capabilities. In summary, LM Studio 0.4.0 significantly enhances functionality, deployment flexibility, and user experience through its comprehensive updates.