1,000,000+ creator records behind two read endpoints: search the index for 10 credits, read any full record for 1. Sixteen fields, one stable schema.
base: https://api.ugcroster.com/v1 · auth: Bearer rsk_…
Most influencer databases are sold as dashboards: you get seats, saved lists and a CSV export button. If the database is meant to sit under your product (a CRM enrichment step, a matching engine, an internal tool), the export button is the bottleneck. The Roster Creator Data API inverts that: the database is the product and REST is the only interface. Two read endpoints cover it: GET /v1/creators/search over the whole index, and GET /v1/creators/{pk} for one full record.
Authentication is one header (Authorization: Bearer rsk_...) against one base URL, https://api.ugcroster.com/v1. The first request works from a bare curl command; an SDK is never required.
search-then-read.sh
LIVE# 1) Search the index (10 credits)
curl "https://api.ugcroster.com/v1/creators/search" \
-H "Authorization: Bearer rsk_live_..." \
-G -d category=Food -d state=NY -d follower_min=10000 -d limit=25
# 2) Read one full record (1 credit)
curl "https://api.ugcroster.com/v1/creators/1893..." \
-H "Authorization: Bearer rsk_live_..."The two-call pattern: a filtered search to find candidates, then 1-credit reads to hydrate the records your product actually shows.
01
Username, full name, follower count, engagement rate (0–1 ratio), category, niches, bio, extracted email, city, state, country, external URL, verification, age, gender and platform key. One schema, every record.
02
q, niche, category, follower_min/max, engagement_min, has_email, city, state, country, gender, age_min/max and sort, combined freely in one request, paginated up to 100 records per page.
03
A profile read costs 1 credit, so keeping your copy of a record fresh is cheap: on Growth ($199/mo, 150,000 credits) that is up to 150,000 full-record reads a month.
Every creator resolves to the same 16-field record. Build your tables once; the schema is owned and versioned by the company that runs the marketplace.
Each 10-credit search answers a structured question (niche, audience band, engagement floor, geography, contact availability) and returns up to 100 records with a pagination object.
Store the pk, then re-read records at 1 credit each when your product renders them or your sync job runs. Rate limits are 60/120/300 requests per minute depending on tier.
Seat pricing assumes a human is looking at the data. When the consumer is code, the honest meter is volume, which is why this API is priced in credits: 10 for a search, 1 for a read, on tiers of 25,000 ($49/mo), 150,000 ($199/mo) and 500,000 ($499/mo) credits, with custom volume above that. A concrete example: the Growth tier’s 150,000 credits fund 15,000 searches a month, and since each search returns up to 100 records, that is up to 1.5 million creator rows flowing into your system monthly for $199.
The other economic question with any influencer database is freshness and provenance. This directory sits behind the UGC Roster marketplace: verified at 1,000,000+ rows and maintained because the marketplace itself depends on it. Fields like extracted_email are filled where available rather than padded, and engagement_rate is computed and stored as a ratio you can filter and sort on.
Because the read side and the write side share one API, the database is also a pipeline: the 15 endpoint groups include briefs, campaigns, applications, contracts, deliverables, messages, payouts, shipments, content tracking, affiliates, analytics and webhooks on brand keys. Your matching engine can find a creator on Monday and your automation can brief, contract and pay the same creator on Tuesday without changing base URLs.
Over 1,000,000 Instagram and TikTok creators, each resolved to one structured record with 16 fields. The count is verified against the live directory that powers the UGC Roster marketplace.
Username, full name, follower count, engagement rate (0–1 ratio), category, niches, biography, extracted contact email where available, city, state, country, external URL, verification status, age, gender and platform key.
Access is metered per call rather than sold as a dump: searches cost 10 credits and return up to 100 records per page, profile reads cost 1 credit. Credit weighting exists precisely so one tier cannot paginate the entire index; for custom volume, talk to the team.
The directory is maintained continuously by the team that operates the marketplace on top of it: the records are production data the marketplace itself runs on. Re-reading a record costs 1 credit, so keeping your copy current is inexpensive.
No. Plans start at $49/month for 25,000 credits (2,500 searches or 25,000 reads). There is no per-seat pricing and no sales call for the self-serve tiers.
Yes. Sitting under someone else’s product is the design goal. One Bearer token, JSON in and out, flat monthly credits, and rate limits of 60–300 requests/minute depending on tier.
Self-serve keys, credit pricing, no sales call. The docs get you from zero to a filtered creator search in one curl command.