How do I install lm studio and verify setup?

# Step 1: Install LM Studio 1. Check system requirements: - **macOS**: Apple Silicon (M1+), macOS 14+, 16GB RAM recommended. - **Windows**: x64/ARM64, 16GB RAM, 4GB VRAM GPU recommended. - **Linux**: x64/ARM64, Ubuntu 20.04+. 2. Download latest version (0.4.6) from [lmstudio.ai/download](https://lmstudio.ai/download). - macOS: .dmg file. - Windows: .exe installer. - Linux: AppImage or curl install script. 3. Install and launch the app. 4. Join Discord for support: [discord.gg/aPQfnNkxGC](https://discord.gg/aPQfnNkxGC).

How do I discover and download your first model?

# Step 2: Download Models 1. Open **Discover** tab. 2. Search for beginner-friendly models like "Phi-3 mini GGUF" or "Gemma-2B Q4_K_M GGUF" (quantized for efficiency). 3. Select a model from Hugging Face integration, check size (start <4GB), and download. 4. Models saved in **My Models** tab. **Tip**: GGUF format for llama.cpp; choose Q4/Q5 quantization for balance of speed/quality.

How do I load model and start chatting?

# Step 3: Basic Chatting 1. Switch to **Chat** tab. 2. Press **Cmd+L (macOS) / Ctrl+L (Win/Linux)** to open model loader. 3. Select downloaded model, choose default load params (e.g., GPU offload if available). 4. Click **Load**; wait for 'Model loaded'. 5. Type prompts and chat. Use **User/Developer mode** for advanced edits (Ctrl+Shift+R). **Tip**: Monitor VRAM/RAM usage in loader.

How do I customize presets and inference parameters?

# Step 4: Tune Presets 1. In Chat tab, select/create **Preset** from dropdown. 2. Edit **system prompt** (e.g., "You are a helpful assistant."). 3. Adjust params: Temperature (0.7 creative, 0.2 precise), Top P (0.9), Max Tokens. 4. **Push** changes or save as new preset. 5. Test in chat; use per-model defaults via gear icon in My Models. See [docs on presets](https://lmstudio.ai/docs/app/presets).

How do I explore rag, server, and cli features?

# Step 5: Advanced Usage 1. **RAG**: Attach .pdf/.docx/.txt files to chat for context-aware queries. 2. **Local Server**: Load model, start **Local Inference Server** tab for OpenAI-compatible API (http://localhost:1234). 3. **CLI**: Run `lms chat` for terminal chats, `lms ls` list models, `lms server start`. 4. Experiment: Query docs, connect apps via API. **Tip**: Use MCP for tools/extensions (docs).

Master LM Studio: Run Local LLMs Privately in 5 Steps

Q: How do I load model and start chatting?

# Step 3: Basic Chatting 1. Switch to **Chat** tab. 2. Press **Cmd+L (macOS) / Ctrl+L (Win/Linux)** to open model loader. 3. Select downloaded model, choose default load params (e.g., GPU offload if available). 4. Click **Load**; wait for 'Model loaded'. 5. Type prompts and chat. Use **User/Developer mode** for advanced edits (Ctrl+Shift+R). **Tip**: Monitor VRAM/RAM usage in loader.

Q: How do I customize presets and inference parameters?

# Step 4: Tune Presets 1. In Chat tab, select/create **Preset** from dropdown. 2. Edit **system prompt** (e.g., "You are a helpful assistant."). 3. Adjust params: Temperature (0.7 creative, 0.2 precise), Top P (0.9), Max Tokens. 4. **Push** changes or save as new preset. 5. Test in chat; use per-model defaults via gear icon in My Models. See [docs on presets](https://lmstudio.ai/docs/app/presets).

Q: How do I explore rag, server, and cli features?

# Step 5: Advanced Usage 1. **RAG**: Attach .pdf/.docx/.txt files to chat for context-aware queries. 2. **Local Server**: Load model, start **Local Inference Server** tab for OpenAI-compatible API (http://localhost:1234). 3. **CLI**: Run `lms chat` for terminal chats, `lms ls` list models, `lms server start`. 4. Experiment: Query docs, connect apps via API. **Tip**: Use MCP for tools/extensions (docs).

🧰

Instant Toolkit

2 artifacts

📋
Step-by-Step Guide

1

Step 1: Install LM Studio

Check system requirements:
- macOS: Apple Silicon (M1+), macOS 14+, 16GB RAM recommended.
- Windows: x64/ARM64, 16GB RAM, 4GB VRAM GPU recommended.
- Linux: x64/ARM64, Ubuntu 20.04+.
Download latest version (0.4.6) from lmstudio.ai/download.
- macOS: .dmg file.
- Windows: .exe installer.
- Linux: AppImage or curl install script.
Install and launch the app.
Join Discord for support: discord.gg/aPQfnNkxGC.

Why this step matters:

-Ensures your hardware supports local AI for smooth performance
-Sets up a private AI environment without cloud dependency

15-30 minutes

Web browser, Computer meeting reqs, LM Studio installer

$0

Definition of Done

•LM Studio app launches without errors
•You can access Discover and Chat tabs

Common Mistakes to Avoid

❌ Installing on unsupported hardware like Intel Mac

→Verify reqs at lmstudio.ai/docs/system-requirements and use compatible machine

❌ Skipping RAM check leading to crashes

→Start with small models (<4GB) on 8GB systems

2

Following along, or just reading? 👀

Spin up a personalized “learn LM Studio” plan you can save, check off, and return to anytime — unlimited on the free trial.

Start free trial →

3

4

5

🎥
Suggested Videos

How to Use LM Studio: A Step-by-Step Guide

Run ANY Open-Source Model LOCALLY (LM Studio Tutorial)

Getting started with LM Studio (100% private local AI)

Local AI Setup That Actually Works - No GPU Required

📚
Learning Resources

📖 Online Courses

Local LLMs via Ollama & LM Studio - The Practical Guide

beginner

Paid

Udemy

LM Studio Course: Run LLMs locally

beginner

Paid

Eduonix

📖 Recommended Books

LM Studio In Action: Building Safe, Private AI

intermediate

by Unknown

🌐 Helpful Websites

LM Studio Official App Docs

documentation

Complete guides for installation, chatting, presets

System Requirements

reference

Detailed hardware specs

LM Studio Discord Community

community

Ask questions and get support

LM Studio GitHub

repository

SDKs and CLI source

🛠️ Useful Tools

LM Studio

Local AI

Desktop app for discovering, downloading, and running local LLMs

lms CLI

CLI

Command-line tool for chatting, server control, model management

Master LM Studio: Run Local LLMs Privately in 5 Steps