In-product feeds
Power reading lists, resource centers, dashboards, or customer portals with semantic feed results.
Add semantic content discovery to internal tools, customer-facing pages, newsletters, or research workflows. The current public API is simple: send semantic queries, receive normalized items.
curl --get "https://curaft.com/api/v1/items" \
--data-urlencode "query=AI search" \
--data-urlencode "query=B2B marketing strategy" \
--data "sort=relevance" \
--data "limit=10"{
"data": [
{
"id": 25083462,
"title": "Ontario to Extend Alcohol Sale and Serve Hours",
"summary": "A concise normalized summary...",
"image": "https://example.com/image.jpg",
"source": {
"name": "Foodservice & Hospitality",
"favicon": "https://www.google.com/s2/favicons?domain=example.com",
"url": "/sources/123-foodservice-hospitality"
},
"published_at": "2026-02-12T09:00:00Z",
"story_url": "/stories/25083462-ontario-extend-alcohol-sale-serve-hours"
}
],
"meta": {
"page": 1,
"limit": 10,
"next_page": 2,
"total": 148,
"source_count": 32
},
"links": {
"self": "https://curaft.com/api/v1/items?query=AI%20search&sort=relevance&limit=10",
"next": "https://curaft.com/api/v1/items?query=AI%20search&sort=relevance&limit=10&page=2"
}
}Send one query or repeat the query parameter for several saved searches. curaft returns normalized items with story URLs, source details, images, timestamps, and pagination metadata.
Power reading lists, resource centers, dashboards, or customer portals with semantic feed results.
Pull fresh stories into brief builders, Slack workflows, internal digests, or lightweight analysis tools.
Keep curaft as the discovery layer while your product owns filtering, layout, alerts, and user permissions.
The current API is rate-limited by IP and works well for prototypes, embeds, and internal tests. Paid semantic item search is available at $5 per 1,000 requests.
Public API
$0
For local tests, proof-of-concept embeds, and checking whether semantic results match your workflow.
Commercial API
$5
For apps, customer portals, newsletters, and internal systems that will need reliable higher-volume access.
The endpoint accepts semantic language directly. Repeat query when you want results from several saved searches. The rendered OpenAPI reference is available at /developers/docs.
querySemantic query
Use one query for an immediate search, or repeat query for each saved search. Example: "AI visibility for B2B SaaS companies".
pagePagination
Start at 1. The response includes next_page and a next link when more results are available.
limitResult count
Return between 1 and 50 items per page. Use smaller limits for embeds and page-speed sensitive UI.
sortResult ordering
Use relevance for best semantic matches, newest for fresh coverage, or oldest for chronological backfills.
time_rangeFreshness window
Use values like last_7_days or last_30_days when you want recent stories only.
Start with one plain-language search, save the useful feeds, and keep the references worth returning to. No sales call, no setup project.