<!-- Markdown twin of https://pixelvault.dev/docs/chatgpt/ -->

# ChatGPT (Custom GPT Action)

Add PixelVault to a ChatGPT [Custom GPT](https://help.openai.com/en/articles/8554397-creating-a-gpt) as an Action, so the GPT can host and manage images. Import this trimmed OpenAPI spec — built specifically for GPT Actions (single server, JSON only):

```
https://pixelvault.dev/openapi-actions.json
```

## Set up the Action

1.  In the GPT editor, open **Configure → Actions → Import from URL** and paste the URL above.
2.  Set **Authentication → API Key** and choose **Bearer**.
3.  Paste a `pv_live_` key from your [dashboard](/dashboard/).

## Available actions

Action

Description

[`upload_image`](/docs/api/images/#upload_image)

Import a public image by its `url` → CDN URL with no fixed expiry (free-plan retention applies)

[`list_images`](/docs/api/images/#list_images)

List your images; search their AI descriptions (`q`) or filter by AI `tag`

[`get_image`](/docs/api/images/#get_image)

Get metadata + CDN URL for one image

[`list_image_tags`](/docs/api/images/#list_image_tags)

The AI tags in your project with how many images carry each

[`describe_image`](/docs/api/images/#describe_image)

Start the AI description (alt text, OCR, tags) of an image you already uploaded

[`delete_image`](/docs/api/images/#delete_image)

Delete one image

GPT Actions can't send file uploads, so `upload_image` takes a public image **URL** rather than a file. Once hosted, apply [transforms](/docs/transforms/) by appending query params to the returned URL. Need base64 uploads, video, audio, or the full tool set? Use the [MCP server](/docs/mcp/) instead: its OAuth endpoint works as a ChatGPT app connector.

## Next steps

-   [MCP server](/docs/mcp/) for the complete tool inventory.
-   [AI descriptions & search](/docs/guides/ai-descriptions/) to understand `describe_image` and `q`.
