Local Studio
A local-first control panel for running and managing self-hosted LLM backends such as vLLM, SGLang, llama.cpp and ExLlamaV3, with GPU monitoring, an OpenAI-compatible chat proxy, and agent sessions against local or remote controllers.
Local Studio is a local-first workstation for running, managing and using self-hosted LLM backends. One machine can launch models, watch GPU and runtime state, chat with OpenAI-compatible endpoints, and run agent sessions against local or remote controllers, with the app organized around Status, Workbench, Configure and Usage surfaces.
It is built from two modules sharing one controller API: a Bun/Hono backend that owns model lifecycle, launching, evicting, recipes, downloads and runtime process coordination, an OpenAI-compatible proxy for chat, models, tokenization and audio, system state such as GPU metrics, logs and settings over server-sent events, and integrations; and a Next.js and React frontend, including a macOS Electron desktop shell, that hosts the agent workbench, a consolidated configuration surface, settings, usage and logs. A companion mobile app pairs with a QR code so the same agent sessions and streaming content are reachable from a phone or tablet.
It requires Bun, Node.js, Python and Git, uses NVIDIA CUDA or Apple Silicon's MLX backend for serving, and installs from source or as a signed macOS build.