# QuickStart (/desktop/quickstart)
Get up and running with Jan in minutes. This guide will help you install Jan, download a model, and start chatting immediately.
### Install Jan [#install-jan]
1. Download Jan
2. Install the app ([Mac](/desktop/install/mac), [Windows](/desktop/install/windows), [Linux](/desktop/install/linux))
3. Launch Jan
### Getting Ready [#getting-ready]
Jan automatically downloads its default foundation model on first launch. Once the download completes, you're ready to chat — no setup required.
### Start Chatting [#start-chatting]
Type your message and start chatting. Jan is ready to go.
Try asking:
- "Explain quantum computing in simple terms"
- "Help me write a Python function to sort a list"
- "What are the pros and cons of electric vehicles?"
Jan has built-in web search and fetch tools, so you can ask about the latest news and current events without any extra setup. If you'd rather use a different provider, you can also configure the [Exa](https://exa.ai) MCP server under **Settings > MCP Servers** with your own API key.
### Download More Models [#download-more-models]
The **Hub** is where you browse and download local models. Each model shows whether it fits your
hardware, so you can pick one your machine can run comfortably.
1. Open the **Hub** from the left sidebar
2. Browse or search for a model, then click **Download**
3. Once downloaded, select it from the model selector in any chat
Want help choosing? See [Manage Models](/desktop/manage-models) and the
[local engine guide](/desktop/local-engine/llama-cpp).
## Local or Cloud [#local-or-cloud]
Jan runs models two ways, and you can mix both in the same app:
- **Local models** run entirely on your machine — private, offline, no API key. This is the default.
- **Cloud models** run on a provider's servers (OpenAI, Anthropic, Google, and more). They need an
API key but let you use the largest frontier models without local hardware.
### Connect a Cloud Provider [#connect-a-cloud-provider]
1. Navigate to **Settings** > **Model Providers**
2. Select a provider (e.g. **OpenAI** or **Anthropic**) and paste your API key
3. Select one of its models from the model selector and start chatting
Running your own server or using a provider Jan doesn't list? Connect any OpenAI- or
Anthropic-compatible endpoint — see [Custom Endpoints](/desktop/remote-models/custom-endpoint).
See all [Cloud Providers](/desktop/remote-models/openai) for per-provider setup.
## App Features [#app-features]
Jan is more than a chat interface. Organize your work with Projects, build custom Assistants, run autonomous Agents, and connect external tools via MCP.
}>Hand Jan a task it works through on its own — in a sandbox or directly on your project.
}>Organize conversations with shared instructions and files.
}>Set up a personal AI assistant tailored to how you work.
}>Autonomous AI that reads files, manages calendars, and takes actions.
}>Extend your AI with web search, code execution, databases, and more.
## Models [#models]
Beyond the default model, Jan comes with a range of foundation models for different tasks — from image understanding and reasoning to coding.
}>Optimized for code generation and reasoning tasks.
}>General purpose instruction-following model.
}>Vision-language model with strong reasoning capability for image understanding.