Generated Media and Base64 Imports
This guide matters because many modern publishing workflows do not start with a file already hosted somewhere. They start with an asset generated by ChatGPT, Claude, Gemini, an internal tool, or another automation step.
What Postly can ingest
Postly can accept media from several common sources:
- base64 payloads
- data URLs
- direct file uploads
- public URLs
- inline media imports during post creation or update
Why base64 is valuable
Base64 is useful when your system already has the media in memory. Instead of forcing a separate hosting step first, the workflow can send the media directly into Postly.
That is especially helpful for:
- AI-generated images
- AI-generated videos
- server-side automation pipelines
- CMS plugins with generated assets
- agent workflows that already have the file bytes available
Best pattern for AI clients
If ChatGPT, Claude, or another client generates the media and still has access to it, the strongest pattern is:
- generate the asset
- keep the same turn or request flow active
- pass the media directly into Postly
- create or update the post immediately
This avoids the common failure mode where the user is asked to download, re-upload, or host the file somewhere else before publishing.
What this does not mean
It does not mean every temporary reference lasts forever. If the client loses access to the asset between steps, Postly cannot recover media it never received. That is why same-flow ingestion matters.
If the asset already has a stable public URL, that can also be a clean ingestion path. Use the format that matches the workflow you already have.
API vs MCP
Use API when
your backend, app, or automation system owns the media and the publish action.
Use MCP when
an AI assistant owns the conversation and needs scoped publishing actions inside the same governed workflow.