> ## Documentation Index
> Fetch the complete documentation index at: https://larkup.de/documentation/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Start your first AI conversation in minutes.

Larkup gets out of your way. There is no server to configure, no pipeline to build. Three steps and you are chatting with your data.

<Steps>
  <Step title="Enter your API key">
    When you open Larkup for the first time you will see the onboarding screen. Paste your API key from any supported provider, including OpenAI, Anthropic, Google, and DeepSeek, then click **Save**. The key is stored locally in your browser and used for chat and embedding calls.

    <Frame caption="Choose a provider and save your API key">
      <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/onboard-screen.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=29a5eb655deff6f7a7461b35c08df50b" alt="Larkup onboarding screen with AI provider settings" width="3012" height="1434" data-path="documentation/images/onboard-screen.png" />
    </Frame>

    <Note>
      You can change your model provider at any time from **Settings → AI Models**.
    </Note>
  </Step>

  <Step title="Go to the Data page and load something">
    Click **Data** in the left sidebar, then choose the source that matches your content.

    <Tabs>
      <Tab title="Website">
        Paste a page or domain URL. You can crawl it directly or search for pages by keyword.

        <Frame caption="Add content from a website">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/website-tab.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=0d40442a5860437f92fee81984e024bd" alt="Website source options in the Data page" width="3012" height="1442" data-path="documentation/images/add/website-tab.png" />
        </Frame>
      </Tab>

      <Tab title="Text">
        Paste notes, transcripts, or any text you want to search, with a title.

        <Frame caption="Adding raw text directly">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/text-tab.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=a248b348f08b4d827a9d48877a36a635" alt="Add Text tab with a title and content field" width="3012" height="1442" data-path="documentation/images/add/text-tab.png" />
        </Frame>
      </Tab>

      <Tab title="Integrations">
        Open **Integrations**, choose Notion or another source, then select **Configure** or **Connect**. After authorization, Larkup adds the selected content to your knowledge base.

        <Frame caption="Choose Notion or another connected data source">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/integrations-tab.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=6294d368e5e1e6c33371ae4ae04d1d15" alt="Data integrations including Notion, Google Drive, Slack, and Jira" width="3012" height="1442" data-path="documentation/images/add/integrations-tab.png" />
        </Frame>
      </Tab>

      <Tab title="CSV">
        Open **Files** and upload a CSV or Excel file. Larkup previews the columns and rows before indexing.

        <Frame caption="Review structured data before indexing">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/upload-excel-data.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=a741fd7b39ebc34cf8e50620bae7de0e" alt="CSV and Excel preview in the Files tab" width="2982" height="1424" data-path="documentation/images/add/upload-excel-data.png" />
        </Frame>
      </Tab>

      <Tab title="PDF">
        Open **Files** and drop in a PDF. You can add one file or stage several files together.

        <Frame caption="Add a PDF from the Files tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/files-tab.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=914db56d80ef8a413e1b9301b991621c" alt="PDF staged for upload in the Data page" width="3012" height="1442" data-path="documentation/images/add/files-tab.png" />
        </Frame>
      </Tab>

      <Tab title="Video">
        Open **Media** to upload a video or import a YouTube URL. Review the duration and estimated processing cost before you confirm.

        <Frame caption="Review a YouTube video before import">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/upload-video-data-page.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=3111ea76189c695938d8c962fd021fb8" alt="YouTube video ready to import from the Media tab" width="2982" height="1424" data-path="documentation/images/add/upload-video-data-page.png" />
        </Frame>
      </Tab>
    </Tabs>

    Once you confirm, Larkup indexes the content in the background. You can watch progress in real time. There is no separate indexing step.

    <Frame caption="Indexing progress for a PDF upload">
      <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/add/indexing-pdf-progress.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=cb933af2eef006bb39a63eae430b3fd2" alt="PDF indexing progress" width="3012" height="1434" data-path="documentation/images/add/indexing-pdf-progress.png" />
    </Frame>
  </Step>

  <Step title="Open Chat and start asking">
    Click **Chat** in the sidebar. Select an LLM from the dropdown at the top and ask anything about your data. Answers are grounded in what you loaded, with source citations shown in the sidebar.

    <Frame caption="Chat answering a question from an uploaded spreadsheet">
      <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/chat/chat-with-csv-chart.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=dd232a9d8bc3ffa59146f72928d0ca0c" alt="Chat generating a bar chart from a CSV upload" width="3012" height="1418" data-path="documentation/images/chat/chat-with-csv-chart.png" />
    </Frame>
  </Step>

  <Step title="OR: Talk to it from social media">
    Connect the same Agent to Telegram, Slack, or Discord and answer questions right where your team already talks.

    <Tabs>
      <Tab title="Telegram">
        <Frame caption="Chatting with the Agent through Telegram">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/connections/telegram-chat-demo.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=c743c77b3d09b5ca50689ab0a678b5f6" alt="Telegram conversation with the Agent" width="1662" height="946" data-path="documentation/images/connections/telegram-chat-demo.png" />
        </Frame>
      </Tab>

      <Tab title="Slack">
        <Frame caption="Chatting with the Agent in Slack">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/connections/slack-chat-demo.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=8191fea13f08ec5d71b4987162e74425" alt="Slack conversation with the Agent" width="1840" height="855" data-path="documentation/images/connections/slack-chat-demo.png" />
        </Frame>
      </Tab>

      <Tab title="Discord">
        <Frame caption="Chatting with the Agent in Discord">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/connections/discord-chat-demo.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=bf9a2e511583e64ff4dac84a1ab6b614" alt="Discord conversation with the Agent" width="1842" height="854" data-path="documentation/images/connections/discord-chat-demo.png" />
        </Frame>
      </Tab>
    </Tabs>

    See [Talk to Your Custom AI from Social Media](/documentation/documentation/guide/public-webhooks) to connect a channel.
  </Step>

  <Step title="Step 5: (Optional) Launch & Deploy" icon="server" badge="Developer">
    Customize your Agent exactly to your needs from **Settings → Agent Customization**:

    <Tabs>
      <Tab title="Prompt">
        Write the system prompt the Agent follows.

        <Frame caption="Agent Customization: Prompt tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/customize-agent/customize-prompt.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=731928b3a7b4e3fc47af1cd50f8869f9" alt="System prompt editor" width="2982" height="1404" data-path="documentation/images/customize-agent/customize-prompt.png" />
        </Frame>
      </Tab>

      <Tab title="Tools">
        Turn built in tools on or off (e.g. semantic search, CSV analysis, form filling).

        <Frame caption="Agent Customization: Tools tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/customize-agent/customize-tools.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=f3b1fb50093b5a8cc7c3129b8d806309" alt="Built in tools with enable toggles" width="2982" height="1216" data-path="documentation/images/customize-agent/customize-tools.png" />
        </Frame>
      </Tab>

      <Tab title="Skills">
        Add custom Skills to extend the Agent's instructions and resources.

        <Frame caption="Agent Customization: Skills tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/customize-agent/customize-skills1.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=42697fe20f7250586eb24e2673711039" alt="Add an Agent skill dialog" width="2982" height="1282" data-path="documentation/images/customize-agent/customize-skills1.png" />
        </Frame>
      </Tab>

      <Tab title="MCP">
        Connect a remote Model Context Protocol server.

        <Frame caption="Agent Customization: MCP tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/customize-agent/customize-mcp.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=4355cfa39aaa3f76ad5ba154b0c9fb83" alt="MCP connections list" width="2982" height="1282" data-path="documentation/images/customize-agent/customize-mcp.png" />
        </Frame>
      </Tab>

      <Tab title="Sandbox">
        Give your Agent access to a remote computer for code execution.

        <Frame caption="Agent Customization: Sandbox tab">
          <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/customize-agent/customize-sandbox1.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=ae82e2584ee357e46ec57df453451817" alt="Sandbox provider picker" width="2982" height="1384" data-path="documentation/images/customize-agent/customize-sandbox1.png" />
        </Frame>
      </Tab>
    </Tabs>

    Once configured, launch your custom AI Agent server, connect to it via the SDK or REST API, and deploy it to production.

    <Frame caption="Settings → Larkup Server showing the local endpoint and API key">
      <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/deploy/server-on.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=bc7e37149d4b25e569f9202296d92404" alt="Larkup Server settings with a running server and API key" width="3012" height="1434" data-path="documentation/images/deploy/server-on.png" />
    </Frame>

    Your Agent server also exposes a complete machine readable API and interactive Swagger/Scalar reference.

    <Frame caption="The interactive API reference for your custom Larkup Agent server">
      <img src="https://mintcdn.com/larkup/13l_fOT1PJRw-f4o/documentation/images/rag/scalar-api2.png?fit=max&auto=format&n=13l_fOT1PJRw-f4o&q=85&s=f1c624acbe60d505b5c02d3d14bf02a4" alt="Scalar API reference for the Larkup server" width="3012" height="1442" data-path="documentation/images/rag/scalar-api2.png" />
    </Frame>

    See [Agent Runtime API](/documentation/documentation/developer/agent-runtime-api) and [Deployment](/documentation/documentation/developer/deploy) for more details.
  </Step>
</Steps>

## What you can do next

Your data is indexed and your server is already running. A few directions to explore:

* [Chat with a video](/documentation/documentation/guide/video-chat) and get answers with timestamps.
* [Analyze a spreadsheet](/documentation/documentation/guide/data-analysis) with charts and summaries in Chat.
* [Sign or fill a document](/documentation/documentation/guide/document-editor) with a live preview.
* [Talk to your Agent](/documentation/documentation/guide/public-webhooks) from Telegram, Slack, or Discord.
* [Add a chat widget](/documentation/documentation/developer/deploy) to your website.
* [Connect an AI agent](/documentation/documentation/developer/agent-runtime-api) through the SDK or REST API.

<Tip>
  Prefer the terminal? The CLI covers the same workflow. See the [CLI Reference](/documentation/documentation/developer/cli).
</Tip>