lobehub/docs/usage/getting-started/file-upload.mdx

---
title: File Upload & Management
description: >-
  Upload files to conversations and knowledge bases — documents, code, images,
  and more. AI agents analyze and reference your files.
tags:
  - LobeHub
  - File Upload
  - Documents
  - Knowledge Base
  - File Management
---

# File Upload & Management

LobeHub provides comprehensive file upload capabilities, enabling AI agents to access, analyze, and reference documents, images, code, and other files during conversations.

## Upload Methods

<Tabs>
  <Tab title="Drag & Drop">
    Drag files directly from your file explorer into the chat input area or conversation window:

    - Works with single or multiple files
    - Supports any file type
    - Files upload and preview automatically
    - Add context or questions, then send
  </Tab>

  <Tab title="Click to Upload">
    Click the paperclip or attachment icon in the input area:

    - Browse your file system
    - Select one or more files
    - Preview before upload
    - Multi-select support
  </Tab>

  <Tab title="Paste from Clipboard">
    Copy any file or take a screenshot, then press `Ctrl+V` (`Cmd+V` on Mac) in the message input:

    - Screenshots appear instantly
    - Ideal for quick screenshot questions
    - Works with copied images and files
  </Tab>
</Tabs>

## Supported File Types

<Tabs>
  <Tab title="Documents">
    - **PDF** — Portable Document Format
    - **Microsoft Office** — Word (.docx), Excel (.xlsx), PowerPoint (.pptx)
    - **Text files** — .txt, .md, .rtf
    - **HTML** — Web pages

    **Capabilities:** Text extraction, summarization, content search, data extraction from spreadsheets.
  </Tab>

  <Tab title="Code & Data">
    - **Code files** — .js, .ts, .py, .java, .cpp, .go, .rs, and more
    - **Web files** — .html, .css, .json, .xml
    - **Data formats** — .csv, .json, .yaml, .sql
    - **Configuration** — .env, .config, .toml

    **Capabilities:** Code review, bug detection, documentation generation, data parsing, syntax highlighting.
  </Tab>

  <Tab title="Images">
    - **Photos** — .jpg, .jpeg, .png, .webp
    - **Graphics** — .gif, .bmp, .svg
    - **Screenshots** — Any image format

    **Capabilities:** Visual analysis (with vision models), text extraction (OCR), image description, design feedback. See [Vision](/docs/usage/getting-started/vision) for details.
  </Tab>

  <Tab title="Media">
    - **Video** — .mp4, .webm, .mov
    - **Audio** — .mp3, .wav, .m4a, .flac

    **Capabilities:** Playback, transcription (with appropriate tools), content analysis.
  </Tab>
</Tabs>

### File Size Limits

- Maximum file size: typically 50 MB per file
- Recommended: Under 20 MB for best performance
- Large files are automatically chunked for processing

## Using Files in Conversations

### Contextual Conversations

When you upload files to a conversation, the AI agent can reference them directly:

```
You: [uploads API_Documentation.pdf]
     "What are the rate limits for the free tier?"

Agent: According to the API documentation, rate limits are:
       - Free tier: 100 requests/hour
       - Pro tier: 1,000 requests/hour
       - Enterprise: Custom limits
       Source: API_Documentation.pdf (page 12)
```

### Multi-File Analysis

Reference multiple files in a single conversation:

```
"Compare the architectures described in design_v1.pdf and design_v2.pdf"
"Summarize all meeting notes from this folder"
"Find contradictions between doc_a.txt and doc_b.txt"
```

### File-Based Workflows

Create workflows around files:

1. Upload source material
2. Extract information or insights
3. Generate derived content (summaries, reports, translations)
4. Export or iterate on results

## File Viewer

LobeHub includes an integrated file viewer:

- **PDF Viewer** — Page navigation, zoom, text selection, search, fullscreen
- **Code Viewer** — Syntax highlighting, line numbers, search, 50+ languages
- **Image Viewer** — Full-resolution viewing, zoom and pan, gallery mode
- **Office Documents** — Formatted preview preserving layout and styling

Click on any file name or thumbnail in chat to open the viewer.

## Knowledge Base Integration

Files uploaded to conversations are automatically saved to the [Resource Library](/docs/usage/getting-started/resource). You can also organize files into dedicated knowledge bases for persistent, searchable access across all your agents.

See [Resource Library](/docs/usage/getting-started/resource) for the full knowledge base workflow.

## Tips for Effective File Use

<AccordionGroup>
  <Accordion title="Organize Files Logically">
    Use descriptive filenames and organize into themed knowledge bases for better retrieval.
  </Accordion>

  <Accordion title="Include Context Documents">
    Upload README files, glossaries, or context documents to help AI understand domain-specific content.
  </Accordion>

  <Accordion title="Keep Files Updated">
    Replace outdated files with new versions to ensure AI references current information.
  </Accordion>

  <Accordion title="Use Text-Extractable Formats">
    Text-extractable formats (PDF, DOCX, TXT) work better than image-only documents for search and analysis.
  </Accordion>

  <Accordion title="Split Large Documents">
    Very large files may benefit from being split into topical sections for better search performance.
  </Accordion>
</AccordionGroup>

## Privacy & Security

<Callout type={'warning'}>
  Uploaded files may be processed by AI services. Avoid uploading sensitive or confidential
  information unless using a private or self-hosted deployment.
</Callout>

- Files are stored in your LobeHub instance
- Content may be sent to AI providers for analysis
- Knowledge base data persists in local/server database
- Delete files when no longer needed

<Cards>
  <Card href={'/docs/usage/getting-started/resource'} title={'Resource Library'} />

  <Card href={'/docs/usage/getting-started/vision'} title={'Vision & Image Understanding'} />

  <Card href={'/docs/usage/getting-started/agent'} title={'Agent'} />
</Cards>