> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/MatthewSabia1/Joip-Web-App-2/llms.txt
> Use this file to discover all available pages before exploring further.

# Smart Captions

> Generate AI-powered NSFW captions for your images with customizable themes and styling

Smart Captions uses AI to generate contextual NSFW captions for your images, with full control over theme, styling, and text appearance.

## Overview

The Smart Captions tool lets you:

* Upload images or fetch from Reddit
* Choose from 6 caption themes (JOI, Forced-Bi, Beta, CBT, Cuckold, Custom)
* Customize text styling (font, size, color, alignment, shadows)
* Generate, remix, and save captioned images
* Build a history of recent creations

## Creating Smart Captions

<Steps>
  <Step title="Upload or Fetch Image">
    Choose your image source from the left panel:

    <Tabs>
      <Tab title="Upload">
        Add images from your device:

        * **Drag and drop** into the drop zone
        * Click **browse** to select files
        * **Paste from clipboard** (Ctrl/Cmd+V)
        * Click **Paste from clipboard** button

        <Warning>
          Only static images (JPEG, PNG, WebP) are supported. GIFs are blocked.
        </Warning>
      </Tab>

      <Tab title="Reddit">
        Fetch random images from subreddits:

        **Subreddit Mode**

        1. Add subreddits using the chips input (up to 25)
        2. Click **Fetch Random Image**
        3. System fetches from your saved list

        **Search Mode**

        1. Switch to Search tab
        2. Enter keywords (e.g., "blonde goddess")
        3. Click **Search Reddit**
        4. Random image from search results loads

        <Note>
          Reddit images are automatically excluded after use to prevent duplicates.
        </Note>
      </Tab>
    </Tabs>

    **Image Preview**

    * Shows resolution and file size overlay
    * Click **Reset** (X button) to clear and start over
    * Image metadata appears in bottom corner
  </Step>

  <Step title="Select Caption Theme">
    Choose the AI caption style (Theme panel):

    <Tabs>
      <Tab title="JOI / Edging">
        Mommy/Goddess gives stroke commands:

        * Edge counts and denial
        * Making them dumb and drippy
        * Control and teasing themes
      </Tab>

      <Tab title="Forced-Bi">
        "It's not gay if..." scenarios:

        * Manipulation tactics
        * Grindr commands
        * Rationalization of acts
      </Tab>

      <Tab title="Bully / Simp">
        Mean girl energy:

        * Instagram stalking
        * Out of your league
        * Pay up loser themes
      </Tab>

      <Tab title="Pain / CBT">
        Specific pain commands:

        * Toilet seats, rulers, punches
        * Squeezes and counts
        * Detailed physical instructions
      </Tab>

      <Tab title="Cuckold">
        She talks about HIM:

        * The bull and BBC
        * Size comparisons
        * Cleanup duties
      </Tab>

      <Tab title="Custom Prompt">
        Write your own AI instructions:

        * Up to 1000 characters
        * Describe voice, personality, style
        * Control tone and length

        <Note>
          Custom prompts require detailed instructions for best results.
        </Note>
      </Tab>
    </Tabs>
  </Step>

  <Step title="Customize Text Styling">
    Adjust caption appearance (Style panel):

    **Font Settings**

    * **Size**: 12-72px slider
    * **Family**: Inter, Georgia, Arial Black, Impact, Trebuchet, Verdana
    * **Color**: White, Black, Gold, Red, Blue, Green, Purple, Pink
    * **Alignment**: Left or Center

    **Effects**

    * **Background**: Toggle semi-transparent background
      * Opacity slider: 0-100%
      * Improves readability on busy images
    * **Text Shadow**: Toggle drop shadow for depth

    <Note>
      Preview updates live on the canvas as you adjust settings.
    </Note>
  </Step>

  <Step title="Generate Caption">
    Click **Generate** button (bottom right):

    1. **Upload Phase**: Image uploads to server (if not from Reddit)
    2. **AI Generation**: Caption created based on theme (2-5 seconds)
    3. **Rendering**: Text overlaid on canvas with your styling
    4. **History Save**: Result saved to Recent Creations (right panel)

    <Warning>
      Generation costs **10 credits** and requires an active internet connection.
    </Warning>

    **What happens during generation:**

    * Spinner overlay appears on canvas
    * Generate button shows loading state
    * Caption appears when ready
    * Settings panels stay accessible
  </Step>

  <Step title="Review and Export">
    Once generated, you can:

    **Edit Caption**

    * Click caption text area to edit
    * Manual edits update canvas in real-time
    * No credit cost for editing

    **Remix**

    * Click **Remix** to generate new caption
    * Uses same image and theme
    * Costs **10 credits**
    * Previous caption is replaced

    **Copy Caption**

    * Click **Copy Caption** button
    * Copies caption text to clipboard
    * Useful for sharing or saving text separately

    **Download**

    * Click **Download** button
    * Saves as PNG with caption baked in
    * Filename: `smart-caption-{timestamp}.png`

    **Save to Vault**

    * Click **Save to Vault**
    * Stores in Media Vault with metadata
    * Includes source subreddit if from Reddit
  </Step>
</Steps>

## Recent Creations

The right panel shows your recent Smart Caption creations:

**History Features**

* Last 8 creations displayed as thumbnails
* Click thumbnail to restore image, caption, and all settings
* **Remix from history**: Reloads original image (no caption overlay)
* Persists in browser localStorage
* Newest creations appear first

<Note>
  Recent Creations include the raw image (without caption) so you can remix with different themes or styling.
</Note>

## Step Indicator

Top progress bar guides you through the workflow:

1. **Upload** - Add or fetch your image
2. **Theme** - Select caption style
3. **Style** - Customize text appearance
4. **Generate** - Create your captioned image

## Advanced Features

### Reddit Integration

**Subreddit Management**

* Add up to 25 subreddits
* Saved to localStorage for quick access
* Autocomplete suggestions as you type
* Remove individual subreddits
* System validates subreddit existence

**Fetch Behavior**

* Fetches 40 posts per request
* Filters for static images only
* Excludes previously used URLs
* Random selection from results
* Retries once on failure

### Custom Prompts

**Writing Effective Prompts**

* Describe the voice/persona (e.g., "dominant mommy", "bratty goddess")
* Specify tone (teasing, cruel, playful)
* Include content focus (body parts, actions, emotions)
* Mention POV (first person, second person)
* Suggest caption length (short/punchy vs detailed)

**Example Custom Prompt**

```
Write as a cruel goddess who mocks the viewer for being 
weak. Use second person POV. Focus on humiliation and 
denial. Keep it under 100 characters. Playful but mean tone.
```

### Styling Tips

<AccordionGroup>
  <Accordion title="Readability">
    For maximum readability:

    * Use **White text** on dark images
    * Use **Black text** on light images
    * Enable **Background** for busy/complex images
    * Enable **Text Shadow** for extra contrast
    * Set background opacity to 40-60%
  </Accordion>

  <Accordion title="Aesthetic Choices">
    Match style to content:

    * **Impact/Arial Black**: Bold, attention-grabbing (memes)
    * **Georgia**: Elegant, sophisticated (artsy images)
    * **Inter**: Clean, modern (default choice)
    * **Gold color**: Luxurious, dominant energy
    * **Pink/Purple**: Playful, bratty themes
  </Accordion>

  <Accordion title="Layout">
    Text placement strategy:

    * **Left align**: Better for longer captions
    * **Center align**: Better for short captions
    * Bottom placement: Classic meme style
    * Larger font (48-72px): Short, punchy captions
    * Smaller font (18-32px): Detailed, multi-line captions
  </Accordion>
</AccordionGroup>

## Workflow Examples

<Tabs>
  <Tab title="Quick Caption">
    Fastest workflow for rapid creation:

    1. Click **Fetch Random Image** (Reddit)
    2. Select theme from dropdown (JOI default)
    3. Click **Generate**
    4. Click **Download** or **Save to Vault**
    5. Click **Reset** to start next caption

    Time: \~10 seconds per caption
  </Tab>

  <Tab title="Custom Upload">
    For specific images:

    1. Drag your image into upload zone
    2. Review image metadata
    3. Choose theme (or write custom prompt)
    4. Adjust text styling to match image
    5. Generate and review
    6. Edit caption text if needed
    7. Export or save

    Time: \~30-60 seconds per caption
  </Tab>

  <Tab title="Remix Session">
    Exploring variations:

    1. Upload image (or fetch from Reddit)
    2. Generate with first theme (e.g., JOI)
    3. Click **Remix** 2-3 times for variations
    4. Switch to different theme (e.g., Forced-Bi)
    5. Generate and compare
    6. Save best versions to vault

    Time: \~2-3 minutes for 5-6 variations
  </Tab>
</Tabs>

## Limitations

<Warning>
  **File Restrictions**

  * Static images only (JPEG, PNG, WebP)
  * GIFs are not supported
  * Maximum 20MB file size for AI processing
  * Maximum 100MB for upload
</Warning>

<Warning>
  **AI Generation**

  * Requires internet connection
  * 2-5 second generation time
  * API failures fall back to themed captions
  * Custom prompts may produce unexpected results
  * SSRF protection blocks localhost/private IPs
</Warning>

## Troubleshooting

<AccordionGroup>
  <Accordion title="Generation fails">
    **Common causes:**

    * Insufficient credits (need 10 credits)
    * File too large (>20MB)
    * Invalid file format (must be static image)
    * Network connection lost
    * API service temporarily unavailable

    **Solutions:**

    * Purchase more credits
    * Compress image before upload
    * Convert GIFs to static frames
    * Check internet connection
    * Try again in a few minutes
  </Accordion>

  <Accordion title="Fetch from Reddit fails">
    **Common causes:**

    * Invalid subreddit name
    * Subreddit has no images
    * Reddit API rate limit
    * Network timeout

    **Solutions:**

    * Verify subreddit exists and is public
    * Add more varied subreddits
    * Wait 30 seconds and retry
    * Switch to Upload mode temporarily
  </Accordion>

  <Accordion title="Caption text cut off">
    **Causes:**

    * Caption too long for image dimensions
    * Font size too large

    **Solutions:**

    * Reduce font size
    * Manually edit caption to be shorter
    * Use left alignment for long captions
    * Enable text wrapping in styles
  </Accordion>

  <Accordion title="Can't paste from clipboard">
    **Causes:**

    * Browser clipboard permissions denied
    * No image in clipboard

    **Solutions:**

    * Allow clipboard access when prompted
    * Copy an image before pasting
    * Use traditional drag/drop or browse instead
  </Accordion>
</AccordionGroup>

## Integration with Other Features

Smart Captions integrates with:

* **Media Vault**: Save captions directly to your vault
* **Community**: Share captioned images (via Media Vault sharing)
* **Sessions**: Captioned images can be used in custom sessions

## Related Features

* [Sessions](/features/sessions) - Create multimedia slideshows
* [Media Vault](/features/media-vault) - Organize saved captions
* [Community](/features/community) - Share your best captions