Curation API overview
The Curation API gives publishers programmatic access to enriched article, OpenRTB, and audience data. If Prism has scanned your content, this is how you pull the enrichment out: per article, in batch, or as an OpenRTB-ready payload your ad stack can merge directly into bid requests.
What the API is for
Every article Prism scans produces a completed analysis: content classification, audience signals, and contextual attributes. The Curation API exposes that analysis over REST so you can use it anywhere: server-side ad decisioning, direct integration into your Prebid setup, data pipelines, or your own tooling. Full reference documentation, including Auth, Endpoints, Examples, and Limits, lives in the console under Curation API Documentation.
API at a glance
| Property | Value |
|---|---|
| Style | REST |
| Auth | Bearer key, created and revoked from Settings. |
| Format | JSON |
| Batch size | Up to 100 URLs or article IDs per request. |
The V1 contract
The V1 API is built around a small set of guarantees:
- Article enrichment. Retrieve completed article analysis by URL or url_id. For URLs that are not yet indexed, pass on_miss=queue to pre-warm them: the API queues the page for scanning so the next request finds it ready.
- Homepage fallback. Pass fallback=true to cascade from the exact page to the site homepage when an article is not yet indexed, so you always get usable site-level enrichment instead of an empty response.
- OpenRTB-ready output. The response is a slim RTD payload designed to merge directly into OpenRTB 2.x site.content.data and user.data. No transformation layer needed between the API and your bid stream.
- Publisher FPD beta guard. Publisher-level first-party data is exposed only when the account feature flag rtd_publisher_fpd is enabled. Accounts without the flag keep the article-level RTD shape, so enabling the flag never breaks existing consumers.
The StackUp taxonomy
Segments in the response are namespaced by segtax ID, so downstream systems always know which taxonomy a segment belongs to:
| segtax | Taxonomy |
|---|---|
| 502 | StackUp content segments. |
| 4 | IAB Audience Taxonomy segments. |
| 501 | Legacy StackUp audience signals. |
| 600 | Publisher first-party data (private taxonomy, behind the beta guard above). |
Batching and credits
Batch requests accept up to 100 URLs or article IDs in one call. Each item is resolved independently, so one unindexed URL never fails the batch, and each item costs 1 API calls credit against your plan. Track consumption on the Usage & Plan page.