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

# Add a real document

> Make a PDF, document, or URL part of your long-term knowledge.

<div className="course-meta" aria-label="Lesson information">
  <span><Icon icon="book-open" /> Essentials · Lesson 3 / 6</span>
  <span><Icon icon="clock-3" /> \~4 min</span>
  <span><Icon icon="user-round" /> Beginner</span>
</div>

So far you've saved short memories. But real knowledge often lives in longer materials: a report, a spec, a PDF, a web page. Instead of summarizing it all by hand, add the material itself and let Mem search inside it.

## What you'll accomplish

<Check>
  You'll add a real document or URL to Mem and ask a question only that material can answer.
</Check>

## Watch how it's done

<iframe src="https://www.youtube.com/embed/ArbCl0Klf0s?autoplay=0" title="Demo: add a real document" allowfullscreen="true" />

## Your turn

<Steps>
  <Step title="Find a real document you care about">
    Choose a PDF, report, spec, Markdown file, or a public web page you'll actually need again.

    Use your own real material if you can.
  </Step>

  <Step title="Add it to Mem">
    Drag the file into the Timeline input, or paste the URL. You can also import it in the Library view.

    Mem parses the content, splits it into chunks, and indexes it.
  </Step>

  <Step title="Wait for it to become searchable">
    Check its status in the Library view. Once it shows as searchable, you can ask questions about it.
  </Step>

  <Step title="Ask a question only that document can answer">
    Type the question in Timeline.

    If you added an architecture review, ask:

    ```text theme={null}
    What does the architecture review say about API rate limiting?
    ```
  </Step>
</Steps>

## What you should see

<Check>
  Mem answers using the document content, with references to specific pages.
</Check>

<Note>
  Can't see it? Make sure the document finished indexing and isn't still showing a processing state in the Library view.
</Note>

## Why this matters

Documents and memories are searched together. The material stays whole as a source, so its details can be referenced and reused by you and your AI later — without you having to remember or re-explain them.

## Go deeper

Want to understand what the Library can hold and how documents get indexed? Read the [library](https://mem.nowledge.co/docs/library) page in the Nowledge Mem Docs.

## Next lesson

Next you'll learn: keep a genuinely valuable AI conversation.

<Card title="Save an AI conversation →" horizontal href="/essentials/save-ai-conversation" />