spb/tools-spboucher-ai

1{2  "slug": "lm-studio",3  "name": "LM Studio",4  "aliases": ["LMStudio", "lms", "llmster"],5  "vendor": { "name": "Element Labs, Inc.", "slug": "element-labs", "website": "https://lmstudio.ai" },6  "homepage_url": "https://lmstudio.ai/",7  "docs_url": "https://lmstudio.ai/docs/app",8  "pricing_url": "https://lmstudio.ai/pricing",9  "license_url": "https://lmstudio.ai/app-terms",10  "product_type": "application",11  "product_status": "active",12  "summary": "Proprietary desktop app for discovering, downloading and running local LLMs with llama.cpp and MLX engines, a chat UI, an OpenAI-compatible server, CLI, SDKs and a headless daemon",13  "description": "LM Studio is a desktop application from Element Labs for running language models locally on macOS (Apple Silicon), Windows (x64 and ARM64) and Linux (x64). Users search Hugging Face from inside the app, download GGUF models that run through llama.cpp on every platform or MLX models on Apple Silicon, and chat with them in a built-in interface that supports document attachments for offline retrieval and acts as an MCP client for connecting tools. A local server exposes OpenAI-compatible endpoints to other applications on the machine or network, with a native REST API in beta.\n\nBeyond the GUI, the `lms` command line handles downloads, daemon and server control, and llmster is a GUI-less build for servers and CI. Python and TypeScript SDKs load models, generate text and embeddings and build agents against the local runtime. In 2026 the company added LM Studio Bionic, an agent for documents, coding and computer control that runs on local models by default and can call larger hosted open models; this introduced Free, Bionic+ and Pro plans and a cloud component with its own terms. The application itself remains proprietary: the terms grant a license for personal and internal business use and prohibit redistribution or offering it as a service.\n\nFor a Mac-centric workflow LM Studio is the most convenient way to compare GGUF and MLX builds of the same model and to expose one as a local API. Limits include the closed source, dependence on a single vendor's release cadence, and the fact that the cloud features change the privacy picture relative to purely local use.",14  "capabilities": ["Model discovery and download from Hugging Face inside the app", "llama.cpp (GGUF) engine on all platforms and MLX engine on Apple Silicon", "Chat UI with offline document retrieval and MCP tool connections", "OpenAI-compatible local server; REST API in beta", "`lms` CLI and llmster headless daemon", "Python and TypeScript SDKs for generation, embeddings and agents", "Bionic agent for documents, coding and computer control (local-first)"],15  "workflows": ["Compare GGUF and MLX builds of a model on a Mac Studio", "Expose a local model to an IDE or app through the OpenAI-compatible server", "Run llmster on a headless machine and manage it with `lms`"],16  "advantages": ["Friendly GUI for non-specialists", "Native MLX support on Apple Silicon", "SDKs and CLI make it scriptable despite being a GUI product"],17  "limitations": ["Proprietary, closed source", "Cloud and Bionic features carry separate terms and plans", "Single-vendor dependence for engine updates"],18  "prerequisites": ["Apple Silicon Mac, x64/ARM64 Windows or x64 Linux with enough RAM for the model"],19  "audiences": ["Developers and analysts who want local models with a GUI", "Mac users exploring MLX models", "Teams exposing local inference to internal tools"],20  "alternatives": ["ollama", "llama-cpp", "mlx-lm", "vllm"],21  "comparison_notes": "Ollama is the open-source (MIT) equivalent with a CLI-first design and the same llama.cpp engine but no MLX path; LM Studio adds a full GUI, MLX support and a proprietary license. llama.cpp is the engine both wrap. mlx-lm covers MLX inference from Python without a GUI. vLLM is for multi-user GPU serving rather than desktop use.",22  "categories": [{ "slug": "local-inference", "primary": true }, { "slug": "apple-silicon-ml", "label": "MLX engine on Apple Silicon" }],23  "tags": ["llm", "macos", "apple-silicon", "desktop", "api", "gguf", "mlx"],24  "use_cases": ["run-local-models", "serve-models"],25  "operating_systems": ["macos", "windows", "linux"],26  "execution_modes": ["local", "cloud"],27  "interfaces": ["desktop", "cli", "api", "sdk"],28  "deployment_models": ["desktop"],29  "integrations": ["Hugging Face Hub", "llama.cpp", "MLX", "Model Context Protocol"],30  "data_formats": ["GGUF", "MLX safetensors", "PDF and text documents (RAG)"],31  "languages": ["Python", "TypeScript"],32  "source_model": "proprietary",33  "pricing_model": ["freemium", "paid"],34  "pricing_checked_at": "2026-11-02",35  "self_hostable": true,36  "sources": [37    { "url": "https://lmstudio.ai/", "type": "official_site", "title": "LM Studio", "supports": ["name", "summary"], "checked_at": "2026-11-02" },38    { "url": "https://lmstudio.ai/docs/app", "type": "official_docs", "title": "LM Studio app documentation", "supports": ["capabilities", "operating_systems", "interfaces", "languages"], "checked_at": "2026-11-02" },39    { "url": "https://lmstudio.ai/pricing", "type": "pricing_page", "supports": ["pricing_model"], "checked_at": "2026-11-02", "note": "Free, Bionic+ and Pro plans" },40    { "url": "https://lmstudio.ai/app-terms", "type": "license_page", "title": "LM Studio application terms", "supports": ["source_model", "vendor"], "checked_at": "2026-11-02", "note": "Proprietary; license for personal and internal business use" }41  ],42  "confidence": "high",43  "discovery_method": "editorial-research",44  "depth": "deep",45  "editorial_status": "reviewed",46  "visibility": "public",47  "featured": false,48  "author": "batch-A"49}50