Skip to main content
Generate AI slideshows, render them as TikTok photo carousels or Instagram carousels, and post them across 6 platforms - all through the API. The full content creation pipeline for OpenClaw agents and AI-driven social media automation.
Use slideshow endpoints to build slide content, iterate on text/layout, render final images, and publish those images through the posting API. This is the core of the automated content pipeline - your OpenClaw agent or automation script generates slideshows, and the Create Post endpoint distributes them to TikTok, Instagram, YouTube, Pinterest, LinkedIn, and Facebook. All slideshow routes are key-scoped and use the standard Partner API envelope documented in Response Pattern.

​
Scope Rules

  • Workspace keys read and write slideshow rows in their workspace.
  • Personal keys read and write rows where user_id = owner_user_id and workspace_id IS NULL.
  • Cross-scope resources return 404 not_found.

​
How It Works

Think of slideshows as a content object you can refine before posting:
  1. Create a slideshow (generate) with AI/manual config, or copy from TikTok.
  2. Iterate on draft content (get, update, optional regenerate per slide).
  3. Render when ready (render) to produce final slide assets.
  4. Publish those slide image URLs using Create Post with media.type = "slideshow" and media.urls.

​
Typical End-to-End Flow

  1. Create/choose a pack (optional but common): Slideshow Packs
  2. Generate a slideshow: Generate Slideshow
  3. Refine the draft: Get Slideshow, Update Slideshow, Regenerate Slide
  4. Render finalized slides: Render Slideshow
  5. Send rendered slide URLs to posts API:
{
  "caption": "My slideshow post",
  "media": {
    "type": "slideshow",
    "urls": [
      "https://cdn.example.com/slide-1.jpg",
      "https://cdn.example.com/slide-2.jpg"
    ]
  },
  "accounts": [{ "id": "0f4f54d4-8cce-4fb7-8c7b-befbcb8af812" }]
}
For publishing, use final image URLs from the slideshow response. After render, this is typically slides[].rendered_image_url (fallback to image_url if a rendered URL is not present).

​
Endpoint Index

OperationEndpointDocs
Preview TikTok copyPOST /api/partner/v1/slideshows/copy-tiktok/previewPreview TikTok Slideshow Copy
Import TikTok copyPOST /api/partner/v1/slideshows/copy-tiktok/importImport TikTok Slideshow Copy
Generate slideshowPOST /api/partner/v1/slideshows/generateGenerate Slideshow
List slideshowsGET /api/partner/v1/slideshowsList Slideshows
Get slideshowGET /api/partner/v1/slideshows/{slideshowId}Get Slideshow
Update slideshowPATCH /api/partner/v1/slideshows/{slideshowId}Update Slideshow
Delete slideshowDELETE /api/partner/v1/slideshows/{slideshowId}Delete Slideshow
Render slideshowPOST /api/partner/v1/slideshows/{slideshowId}/renderRender Slideshow
Duplicate slideshowPOST /api/partner/v1/slideshows/{slideshowId}/duplicateDuplicate Slideshow
Regenerate one slidePOST /api/partner/v1/slideshows/{slideshowId}/slides/{slideIndex}/regenerateRegenerate Slide

​
Validation Highlights

  • Standard generation and manual setup both support 1-10 slides.
  • Any image_pack slide must resolve a pack through global pack_id or per-slide pack_assignments[index].
  • slide_config index maps must use 0-based numeric keys inside range.
  • Updating slides on a rendered slideshow resets it to draft and clears rendered artifacts.

​
Packs and Templates

​
UI vs API

  • UI is usually the fastest way to curate packs and templates manually.
  • API is better for automation, external pipelines, and agent-driven workflows - especially when using OpenClaw agents or Claude Code to automate TikTok slideshow creation and posting.
  • Both paths use the same underlying slideshow, pack, and template models.

​
Agent Automation Tips

If you’re building an OpenClaw agent or AI automation pipeline, here’s the optimal slideshow workflow:
  1. Batch generate - Create multiple slideshows in sequence, each with different hooks and styles
  2. Render in bulk - Render all slideshows before posting to avoid delays
  3. Post as drafts to TikTok - Use tiktok.post_mode: "MEDIA_UPLOAD" so you can add trending sounds manually
  4. Post directly to other platforms - Instagram, YouTube, Pinterest, LinkedIn, and Facebook support direct posting
  5. Track analytics - Use the Analytics endpoints to see what’s performing and feed that data back to your agent
See our official OpenClaw skill for a ready-made implementation of this workflow.