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

# Upload Your First File

> Attach documents, images, audio, video, code, and data files to any ZeroTwo chat and learn how each file type is processed.

ZeroTwo supports a wide range of file types directly in chat. Attach a file to any message to ask questions about its contents, have the AI summarize or analyze it, extract data, or use it as the basis for a Canvas document.

<Warning>
  Do not upload files containing sensitive personal data such as Social Security numbers, passwords, private keys, or financial account credentials. See [Privacy & Data](/privacy/data-and-compliance) for ZeroTwo's data handling policies.
</Warning>

***

## Two Ways to Upload

### Paperclip Icon

Click the **paperclip icon** in the prompt bar (to the left of the text input) to open your operating system's file picker. You can select multiple files at once using Shift-click or Cmd/Ctrl-click.

### Drag and Drop

Drag one or more files from your desktop or file manager and drop them anywhere in the main chat area. A drop zone highlights to confirm the target. The file thumbnails appear in the prompt bar ready to send.

***

## Step-by-Step

<Steps>
  <Step title="Click the paperclip icon or drag a file into the chat">
    Both methods work in any active chat or a new conversation. The file begins uploading immediately.
  </Step>

  <Step title="A file preview appears above the text input">
    You will see a thumbnail or icon for each attached file in the prompt bar. Click the × on a thumbnail to remove a file before sending.
  </Step>

  <Step title="Type your question (or leave the input blank)">
    Enter your instruction or question about the file. If you leave the prompt blank, ZeroTwo will infer what to do based on the file type — for example, it will summarize a PDF or describe an image.
  </Step>

  <Step title="Send">
    Press **Enter** to send. ZeroTwo processes the file and generates a response. You can continue asking follow-up questions about the file in subsequent turns of the same conversation without re-attaching it.
  </Step>
</Steps>

***

## What ZeroTwo Does with Each File Type

<Tabs>
  <Tab title="Documents">
    **Supported formats:** PDF, DOCX, PPTX, XLSX, and other common document types.

    ZeroTwo performs full text extraction from documents — including text inside tables, headers, and footnotes. For multi-page PDFs, ZeroTwo is aware of page structure and can answer page-specific questions.

    **Things to try:**

    * "Summarize the key findings from this PDF in 5 bullet points."
    * "What does the contract say about termination clauses?"
    * "Extract all the data from the tables in this report."
    * "Rewrite this Word document in a more formal tone."

    <Tip>
      For long PDFs, you can ask ZeroTwo to focus on a specific section — for example: "What does section 3 say about pricing?" Page-level context is preserved.
    </Tip>
  </Tab>

  <Tab title="Images">
    **Supported formats:** PNG, JPG/JPEG, GIF, WEBP, BMP, HEIC, and other common image formats.

    Images are analyzed using the selected model's **vision capability**. ZeroTwo can describe images, extract text via OCR, analyze charts and diagrams, and answer detailed visual questions.

    **Things to try:**

    * "Describe what is happening in this image."
    * "Extract all the text visible in this screenshot."
    * "What trends does this chart show?"
    * "Compare these two product mockups and suggest improvements."

    <Info>
      Vision analysis requires a model with image input support. Models such as GPT-4.1, Claude Sonnet 4.6, and Gemini 2.5 Pro all support vision. The model picker indicates which models support image input.
    </Info>
  </Tab>

  <Tab title="Audio">
    **Supported formats:** MP3, WAV, M4A, AAC, OGG, FLAC, and other common audio formats.

    Audio files are automatically transcribed using **Whisper** before being passed to the model. The full transcript is used to answer questions, generate summaries, or extract structured information.

    **Things to try:**

    * "Transcribe this recording and format it as a meeting summary."
    * "List all the action items mentioned in this audio."
    * "Translate this audio transcript into Spanish."
    * "Identify the key topics discussed in this podcast clip."
  </Tab>

  <Tab title="Video">
    **Supported formats:** MP4, MOV, WEBM, AVI, MKV, and other common video formats. Max size: **200 MB**.

    ZeroTwo processes video by extracting key frames at regular intervals and analyzing them with vision models. This allows ZeroTwo to describe scenes, read on-screen text, and answer questions about visual content.

    **Things to try:**

    * "Describe what happens in this video."
    * "Read the text visible on screen in this recording."
    * "Summarize the content of this tutorial video."
    * "Identify the objects and people in each scene."

    <Note>
      Video processing extracts frames rather than performing full temporal analysis. For best results, use clearly filmed video with distinct, well-lit content at each scene. Fast-motion or highly compressed video may reduce analysis accuracy.
    </Note>
  </Tab>

  <Tab title="Code & Data">
    **Code files:** `.py`, `.js`, `.ts`, `.jsx`, `.tsx`, `.go`, `.rs`, `.java`, `.c`, `.cpp`, `.rb`, `.php`, `.swift`, `.kt`, `.sh`, and other source code extensions.

    **Data files:** `.json`, `.csv`, `.yaml`, `.yml`, `.ipynb` (Jupyter notebooks), and other structured data formats.

    ZeroTwo can read, explain, analyze, and refactor code files, as well as query and summarize structured data.

    **Things to try:**

    * "Review this code for bugs and suggest improvements."
    * "Explain what this function does in plain English."
    * "Refactor this Python script to follow PEP 8 style."
    * "Analyze this CSV and describe the distribution of the 'revenue' column."
    * "Convert this JSON config to YAML format."
  </Tab>
</Tabs>

***

## File Limits

| Limit                        | Value           |
| ---------------------------- | --------------- |
| Max files per message (Pro+) | 10 files        |
| Max files per message (Free) | 1 file          |
| Max file size (most types)   | 50 MB per file  |
| Max file size (video)        | 200 MB per file |
| Upload timeout               | 60 seconds      |

<Warning>
  Files that exceed the size limit or take longer than 60 seconds to upload will fail with an error. For large video files, ensure you have a fast and stable internet connection. Try compressing the video or splitting it into segments if uploads time out.
</Warning>

***

## Files Library

All files you have ever uploaded to ZeroTwo are saved to the **Files Library** at `/files`. From there you can:

* Browse all past uploads sorted by date
* Preview documents and images inline
* Reuse a file in a new chat without re-uploading it
* Delete files you no longer need

To open the Files Library, click **Files** in the left sidebar or topbar, or navigate directly to `/files`.

***

## Tips for Working with Files

<Accordion title="Ask follow-up questions without re-attaching">
  After uploading a file, you can ask as many follow-up questions as you like within the same conversation. The model retains the file context for the entire chat session — no need to re-attach on each message.
</Accordion>

<Accordion title="Combine files with tools">
  Enable **Web Search** alongside a document attachment to cross-reference its contents with current web information. Enable **Canvas Mode** to have ZeroTwo generate a structured document based on your uploaded file.
</Accordion>

<Accordion title="Upload multiple file types in one message">
  You can mix file types in a single upload. For example, attach a PDF report and a CSV dataset together, then ask ZeroTwo to correlate findings between the two.
</Accordion>

<Accordion title="Use Projects to organize files">
  If you regularly work with a set of files (like a codebase or a research collection), add them to a **Project**. Project files are available across all chats within that project without needing to re-attach them. See [Projects](/projects) for details.
</Accordion>

***

## Next Steps

<CardGroup cols={2}>
  <Card title="Files Library" icon="file" href="/files">
    Manage and reuse all your uploaded files from the Files Library.
  </Card>

  <Card title="Canvas Mode" icon="layers" href="/tools/canvas">
    Use uploaded files as the basis for generating rich documents, spreadsheets, and diagrams.
  </Card>

  <Card title="Troubleshooting Uploads" icon="refresh-cw" href="/troubleshooting/upload-errors">
    Diagnose and fix common file upload errors.
  </Card>

  <Card title="Choose a Tool" icon="zap" href="/getting-started/choose-a-tool">
    Combine file uploads with Web Search, Canvas Mode, and more.
  </Card>
</CardGroup>