# lmstudio.ai

> AI-optimized mirror of lmstudio.ai containing 50 pages totalling 48,774 words of clean markdown content, structured data, and semantic HTML. Original source: https://lmstudio.ai/. Last updated: 2026-05-01T13:12:58.676Z.

Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG).

## Homepage

- [Run AI models, locally and privately.](/site-root.html): Run local AI models like gpt-oss, Llama, Gemma, Qwen, and DeepSeek privately on your computer. (146 words)

## Articles & Blog Posts

- [feniks0113/rag-v2/files/examples/lmstudio-mcp-json.html](/feniks0113/rag-v2/files/examples/lmstudio-mcp-json.html) (1 words)
- [models/index.html](/models/index.html) (1 words)
- [blog/unified-mlx-engine/index.html](/blog/unified-mlx-engine/index.html) (1 words)
- [API Changelog](/docs/sitemap-xml.html) (161 words)
- [yongwei/rag-flex/files/readme-md.html](/yongwei/rag-flex/files/readme-md.html) (1 words)
- [blog/0-4-0.html](/blog/0-4-0.html) (1 words)
- [The .act() call](/docs/python/agent/act/index.html): How to use the .act() call to turn LLMs into autonomous agents that can perform tasks on your local machine. (1,262 words)
- [Beledarian's LM Studio Tools](/hgan/beledarians-lm-studio-tools/index.html): This project is a plugin for LM Studio that provides a rich set of tools to the large language model. It allows the LLM to interact with your local file system, execute code, access the internet, and more. This empowers the LLM to perform complex tasks like code generation, project scaffolding, and web research. (1,080 words)
- [big-rag](/mindstudio/big-rag/files/quickstart-md.html): Big RAG Plugin for LM Studio
A powerful RAG (Retrieval-Augmented Generation) plugin for LM Studio that can index and search through gigabytes or even terabytes (not tested) of document data. Hosted here: github.com/ari99/lm_studio_big_rag_plugin
 (814 words)
- [LM Studio 0.3.10: 🔮 Speculative Decoding](/blog/lmstudio-v0-3-10.html): Inference speed up with Speculative Decoding for `llama.cpp` and `MLX` (1,734 words)
- [LM Studio 0.3.15: RTX 50-series GPUs and improved tool use in the API](/blog/lmstudio-v0-3-15.html): Support for CUDA 12, new system prompt editor UI, improved tool use API support, and preview of community presets. (1,050 words)
- [Open Responses](/blog/openresponses/index.html): Update to LM Studio 0.3.39 for Open Responses support (876 words)
- [Fields](/docs/typescript/api-reference/llm-prediction-config-input/index.html): Standalone LM Studio documentation for the app, developer APIs, SDKs, CLI, and integrations. (840 words)
- [Authentication](/docs/developer/core/authentication/index.html): Using API Tokens in LM Studio (358 words)
- [Frequently Asked Questions](/docs/lmlink/basics/faq/index.html): Answers to common questions about LM Link. (664 words)
- [persistent-memory](/khtsly/persistent-memory/files/readme-md.html): Give your local LLM a long-term brain. This plugin stores facts, preferences, projects, and notes across conversations in a local SQLite database — zero cloud, zero external services, fully portable.

https://github.com/imezx/persistent-memory-plugin (298 words)
- [OpenAI-Compatible Endpoint Adapter for LM Studio](/ankh/openai-compat-endpoint/files/readme-md.html): qwen/qwen-turbo

x-ai/grok-4.1-fast:free

openrouter/bert-nebulon-alpha

etc FREE  @openrouter.ai 

baseUrl: "https://openrouter.ai/api/v1" (382 words)
- [LM Studio 0.4.10](/changelog/lmstudio-v0-4-10.html): Release notes for LM Studio 0.4.10 (41 words)
- [Offline Operation](/docs/app/offline/index.html): LM Studio can operate entirely offline, just make sure to get some model files first. (523 words)
- [Hermes Agent](/docs/integrations/hermes/index.html): Use Hermes Agent with LM Studio (176 words)
- [rag-v2](/systemnova/rag-v2/index.html): Custom plugin by systemnova (322 words)
- [Set a preferred device](/docs/lmlink/basics/preferred-device/index.html): Choose a preferred device to load models via LM Link (129 words)
- [LM Studio as a Local LLM API Server](/docs/developer/core/server/index.html): Run an LLM API server on localhost with LM Studio (146 words)
- [OpenClaw](/docs/integrations/openclaw/index.html): Use OpenClaw with LM Studio (213 words)
- [lms import](/docs/cli/local-models/import/index.html): Import a local model file into your LM Studio models directory. (225 words)
- [lms — LM Studio's CLI](/docs/cli/index.html): Get starting with the lms command line utility. (433 words)
- [LM Studio, llmster, and lms](/docs/app/basics/lmstudio-vs-llmster-vs-lms/index.html): Understand the differences between the LM Studio app, llmster, and lms (464 words)
- [Improve `openai/gpt-oss` in-chat tool calling reliability](/blog/lmstudio-v0-3-23.html): Improve in-app chat tool calling reliability for gpt-oss, and ability to place MoE expert weights on CPU (533 words)
- [web-search](/brius/web-search/files/readme-md.html): Mirror of nub235/web-search (632 words)
- [Use MCP Servers](/docs/app/mcp/index.html): Connect MCP servers to LM Studio (351 words)
- [olmOCR 2](/models/olmocr2/index.html): The olmOCR 2 model is a Vision Language Model (VLM) from Allen AI. (241 words)
- [gemma-4-e4b](/models/google/gemma-4-e4b/index.html): Gemma4, effective 4B version. Supports image input, reasoning, and tool calling. (2,013 words)
- [qwen35-9b](/sherckuith/qwen35-9b/index.html): Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility. This is a 9B parameter dense model, supporting a native context length of 262,144 tokens. (206 words)
- [duckduckgo](/danielsig/duckduckgo/index.html): An LM Studio plugin that provides access to DuckDuckGo web search and image search. (233 words)
- [LM Studio Beta Releases](/beta-releases/index.html): Beta and release-candidate versions of LM Studio. (84 words)
- [Get Context Length](/docs/python/model-info/get-context-length/index.html): API to get the maximum context length of a model. (315 words)
- [Stateful Chats](/docs/developer/rest/stateful-chats/index.html): Learn how to maintain conversation context across multiple requests (331 words)
- [hermes-4.3-36b](/nousresearch/hermes-4-3-36b.html): Hermes 4.3 36B is a frontier, hybrid-mode reasoning model based on ByteDance Seed 36B base, made by Nous Research that is aligned to you. (1,033 words)
- [qwen3.6-27b](/models/qwen/qwen3-6-27b.html): Dense 27B Qwen 3.6 prioritizes stability and real-world utility, offering developers a more intuitive, responsive, and genuinely productive coding experience. (1,073 words)
- [nemotron-3-nano-4b](/models/nvidia/nemotron-3-nano-4b/index.html): General purpose reasoning and chat model trained by NVIDIA (164 words)
- [qwen3-1.7b](/models/qwen/qwen3-1-7b.html): The 1.7B parameter version of the Qwen3 model family. (140 words)
- [Claude Code](/docs/integrations/claude-code/index.html): Use Claude Code with LM Studio (221 words)
- [LM Studio 0.4.2](/changelog/lmstudio-v0-4-2.html): Release notes for LM Studio 0.4.2 (162 words)
- [Unload a model](/docs/developer/rest/unload/index.html): Unload a loaded model from memory (71 words)
- [ministral-3-8b](/models/mistralai/ministral-3-8b/index.html): An 8B parameter model in the Ministral 3 family, combining a language model with a vision encoder for efficient edge deployment. (123 words)
- [lms daemon up](/docs/cli/daemon/daemon-up/index.html): Start llmster from the CLI. (124 words)
- [qwen3-4b](/models/qwen/qwen3-4b/index.html): The 4B parameter version of the Qwen3 model family. (144 words)
- [nousresearch](/nousresearch/index.html): nousresearch on the LM Studio Hub (52 words)
- [Tools](/llms-txt.html) (28,186 words)

## Resources

- [Full Page Index](/index.html): Browse all cached pages with rich metadata
- [About This Cache](/content/about.html): Methodology, technical details, and usage guidelines
- [XML Sitemap](/content/sitemap.xml): Machine-readable sitemap for crawler discovery
- [Robots.txt](/content/robots.txt): Crawler directives