All vikis
007 First Light
Aion 2
Ananta
ARC Raiders
ArcheAge Chronicles
Call of Duty: Modern Warfare 4
ChatGPT
Chrono Odyssey
Civilization 3
Civilization 4
Claude
Crimson Desert
Deadlock
Diablo 4
Diablo 5
Enshrouded
Fable
Final Fantasy Resonance
Football Manager 27
Forza Horizon 6
God of War
Grand Theft Auto 6
Heroes of Might and Magic 3LM Studio
LM Studio is a free desktop app for Mac, Windows and Linux that finds, downloads and runs open AI models such as Gemma, Qwen and gpt-oss on your own computer, free for work use since Jul 2025. Its newer agent app, Bionic, adds paid US-hosted models from $20 a month.
LM Studio is a free desktop app for running open AI models on your own computer: you search for a model, download it from Hugging Face, and chat with it, with no data leaving your machine. It runs on Apple Silicon Macs, Windows and Linux, uses llama.cpp and Apple's MLX underneath, and can serve your models to other apps through an OpenAI-compatible API. Since Jul 8, 2025 it is free for use at work too, with no license or form needed. In Jul 2026 the company added Bionic, a separate agent app for open models, whose cloud plans start at $20 a month (as of Nov 6, 2026).
Pick LM Studio if you want local AI without a terminal: a model browser that shows what fits your machine, a chat window, and settings you can see. Pick Ollama if you live in the terminal, want the lightest background service, or need local models behind coding tools with one command. For models too big for your computer, use a hosted service such as OpenRouter. See Run AI Locally.
Quick facts
| Maker | LM Studio |
| Get it | lmstudio.ai: macOS (Apple Silicon), Windows (x64 and ARM), Linux; Locally app on iPhone and iPad |
| Price | Free, including for work (since Jul 8, 2025); Bionic+ $20 a month, Pro $100 (as of Nov 6, 2026) |
| Latest major version | LM Studio 0.4.0 (Jan 28, 2026) |
| Engines | llama.cpp (GGUF) and MLX on Apple Silicon |
| Models | Any GGUF or MLX model on Hugging Face: gpt-oss, Gemma, Qwen, Llama, DeepSeek and more |
| Best for | Running local models with a desktop interface |
Hardware: will it run on my computer?
| System | Requirement |
|---|---|
| Mac | Apple Silicon (M1 or later), macOS 14 or later; Intel Macs not supported. 16 GB of memory recommended; 8 GB works with small models |
| Windows | x64 with AVX2, or ARM (Snapdragon X Elite); 16 GB of memory and at least 4 GB of dedicated graphics memory recommended |
| Linux | x64 or ARM64, as an AppImage, Ubuntu 20.04 or later |
Source: LM Studio system requirements, read Nov 6, 2026. As a rule, a model needs a little more memory than its download size, plus room for the conversation. The Mac's unified memory lets it hold larger models than a PC with the same graphics card memory.
Features
Model search and download. LM Studio searches Hugging Face from inside the app and downloads GGUF and MLX builds of models such as gpt-oss, Llama and Qwen, per LM Studio's docs.
Chat with your documents. Attach files to a chat and ask about them offline; LM Studio retrieves the relevant passages for the model.
Local server and API. LM Studio serves models through OpenAI-compatible endpoints on your computer or network, plus its own REST API. Version 0.4.0 added a stateful /v1/chat endpoint, parallel requests to the same model and llmster, a headless version for servers with no window. Since Jan 2026 its models also work in Claude Code (LM Studio blog, Jan 30, 2026).
MCP and developer tools. Local models can use MCP servers (tools such as file access or search), and developers get Python and TypeScript SDKs and the lms command line (docs).
LM Link and the iPhone app. LM Link connects your devices with end-to-end encryption, so the large model on your desktop can answer from your phone; since Jun 4, 2026 that works on iPhone and iPad through Locally, an app whose makers joined LM Studio in Apr 2026.
Bionic. Released Jul 16, 2026, Bionic is a separate app, an agent that codes, researches and edits documents, spreadsheets and slides, with offline voice typing through Mistral's Voxtral. It runs local models through the LM Studio engine, or open models hosted by LM Studio in the US, which LM Studio says keep zero data.
Plans and prices
| Plan | Price | What you get |
|---|---|---|
| Free | $0 | LM Studio, local models, the Bionic agent, LM Link for up to 5 devices, limited web search |
| Bionic+ | $20 a month | US-hosted open models (Kimi K3, GLM 5.3, DeepSeek V4 Flash and others), discounted tokens, fuller web search |
| Pro | $100 a month | 5 times Bionic+ usage, early access to features |
| Enterprise | Contact sales | SSO, control over which models and MCPs staff use |
Prices as of Nov 6, 2026. Team subscriptions are listed as coming soon.
LM Studio vs Ollama
Both are free and run the same open models with the same llama.cpp engine at heart, and both offer an OpenAI-compatible local API. LM Studio is a full desktop app: a model browser that filters by your hardware, detailed settings, document chat, MCP tools and LM Link to your phone. Ollama is open source under MIT, lighter, driven from the terminal with a simpler chat app, and its ollama launch sets up coding agents in one step.
Other alternatives
- llama.cpp: the open-source engine under both, for full control; see Run AI Locally.
- Hugging Face: the hub LM Studio downloads from, with hosted inference when a model will not fit.
- OpenRouter: one API for hosted models, open and closed.
Frequently asked questions
Is LM Studio free? Yes. LM Studio is free for personal and work use since Jul 8, 2025; only the optional Bionic cloud plans cost money, from $20 a month, as of Nov 6, 2026.
Is LM Studio safe and private? Models you run in LM Studio stay on your computer and work offline; chats sent to LM Studio's hosted models are, LM Studio says, kept under zero data retention.
Does LM Studio work on Intel Macs? No. LM Studio needs an Apple Silicon Mac (M1 or later) with macOS 14 or newer.
LM Studio or Ollama: which is better? LM Studio suits people who want a desktop app with a model browser and visible settings; Ollama suits terminal users and open-source fans who want a light background service. Both are free.
What is LM Studio Bionic? Bionic, released Jul 16, 2026, is LM Studio's separate agent app for open models, for coding, research and document work, running local models free or hosted ones on paid plans.
Sources
- LM Studio, read Nov 6, 2026
- Pricing, LM Studio, read Nov 6, 2026
- System requirements, LM Studio docs, read Nov 6, 2026
- LM Studio docs, read Nov 6, 2026
- LM Studio is free for use at work, LM Studio, Jul 8, 2025
- Introducing LM Studio 0.4.0, LM Studio, Jan 28, 2026
- Use your LM Studio Models in Claude Code, LM Studio, Jan 30, 2026
- Locally AI joins LM Studio, LM Studio, Apr 8, 2026
- Run (your largest) local models from your iPhone, LM Studio, Jun 4, 2026
- Introducing LM Studio Bionic, LM Studio, Jul 16, 2026
- LM Studio Full Tutorial - The BEST Way to Run Local Models, Tech With Tim, Sep 26, 2026 (video)
Moments
LM Studio Full Tutorial - The BEST Way to Run Local Models
Tech With Tim
- LM Studio is a free desktop app for running open AI models on your own computer: you search for a model, download it…
- Pick LM Studio if you want local AI without a terminal: a model browser that shows what fits your machine, a chat…
- Model search and download. LM Studio searches Hugging Face from inside the app and downloads GGUF and MLX builds of…
- Chat with your documents. Attach files to a chat and ask about them offline; LM Studio retrieves the relevant passages…
- Local server and API. LM Studio serves models through OpenAI-compatible endpoints on your computer or network, plus…
Photos
Related
-
-
-
-
-
CometComet is Perplexity's AI web browser, free for everyone since Oct 2, 2025 and available on Mac, Windows, iOS and Android. Its built-in Comet Assistant can read pages and carry out tasks for you; the agent needs a paid Perplexity plan.
-
Gemini in ChromeGemini in Chrome is Google's Gemini assistant built into the Chrome browser on Windows, Mac, Chromebook Plus and Android. It reads your open tabs and connects to Google apps; its agent, auto browse, is rolling out to Google AI Pro and Ultra subscribers, first in the US.
-
-
Character.AICharacter.AI is a platform for chatting and role-playing with AI characters, free with ads and optional plans ((c.ai) lite from $4.99 a month, or c.ai+). It filed for Chapter 11 on Oct 12, 2026 to sell itself, with Disney as the lead bidder; the app still runs, and users under 18 cannot use open-ended chat.
See also
-
-
-
Run AI LocallyTo run AI on your own computer, install Ollama or LM Studio (both free) and download an open model sized to your memory, such as Gemma 4, Qwen 3.5 or 3.8, or gpt-oss 20b. With 8 to 16 GB a small model runs well; a 24 GB graphics card or a 32 GB Mac runs 27B to 31B models.
-
Claude CodeClaude Code is Anthropic's agentic coding tool for the terminal, VS Code, JetBrains, a desktop app and the web. It comes with Claude Pro ($20 a month), Max ($100), Ultra ($200), Team and Enterprise, or is billed per token through the Claude API; the free Claude plan does not include it.
-
-
GemmaGemma is Google's family of free open-weight models, built from the same research as Gemini. The current generation, Gemma 4 (Apr 2026), comes in E2B, E4B, 12B, 26B and 31B sizes under the Apache 2.0 license and runs on phones, laptops and single GPUs.