Subscription figures are provider list prices recorded in treg.to’s own catalog grid; per-call prices are what treg.to charges today, with $0.000 added.
set up treg — https://treg.to/llms.txt
Using treg, find 20 TikTok creators posting about home espresso with between 50k and 500k followers. Show me the price first, then give me handles, follower counts and bios.
Each platform has its own index. "Creators" alone makes the agent guess.
These are keyword searches over profiles, not semantic search.
Follower count, bio and link come back; say so and you get a table.
Follower ranges are your filter on the results, not a search parameter.
treg.to holds the provider keys. Neither you nor the agent sees them.
The provider's own rate, $0.000 markup, from a prepaid balance.
Charged per call. $1.00 free per new team, no card to start.
Already pay Hunter? Register it and those calls are never metered.
Another provider is a different word in the prompt, not a new integration.
No SDK, no OAuth dance per vendor, no seats.
treg.to does not choose for you. It hands ChatGPT this comparison, with the price shown before any call, and ChatGPT picks. Or you tell it how: "cheapest", "most reliable", "the one that takes what I have", or a provider by name.
TikHub at $0.001 · TikTok
ScrapeCreators at $0.00188 · Instagram
X (Twitter) at $0.01 · X (Twitter)
Those units are not interchangeable: one call can return many results, so compare on the unit you will actually be billed in.
| Provider | Success | Median | Sample |
|---|---|---|---|
| 100% | 805ms | 15 calls | |
| 100% | 10.2s | 122 calls |
Measured on treg.to traffic: real calls, real inputs, and sample sizes differ by provider. Live reliability, not a controlled benchmark.
| Provider | Price | Accepts | Success rate | Verified |
|---|---|---|---|---|
| $0.00188 per call | query, cursor | 100% (122 calls) | 2026-07-28 | |
| $0.002 per found | keyword, query, rank_token | (3 calls) | 2026-07-28 |
| Provider | Price | Accepts | Success rate | Verified |
|---|---|---|---|---|
| $0.001 per found | keyword, offset, count, user_search_follower_count, user_search_profile_type, user_search_other_pref | 100% (15 calls) | 2026-07-28 | |
| $0.00188 per call | query, cursor, trim | (3 calls) | unverified | |
| $0.0148 per found | keyword, cursor, searchId | (1 calls) | 2026-07-28 |
| Provider | Price | Accepts | Success rate | Verified |
|---|---|---|---|---|
| X (Twitter) | $0.01 per result | query, max_results, next_token | (1 calls) | unverified |
| Provider | Price | Accepts | Success rate | Verified |
|---|---|---|---|---|
| $0.001 per found | channel_id, search_query, language_code, country_code, continuation_token, keyword | (2 calls) | 2026-07-28 |
treg call tikhub.x.tiktok-app-v3-fetch-user-search-result --query keyword=Cat --query offset=0 --query count=5
Swap the id for any provider above. All 10 endpoints behind this job, with their parameters and captured responses, are on the Instagram, TikTok, X (Twitter), YouTube shelf.
Providers return keyword matches and your agent filters them, so ask for more results than you need.
To find creators by what they posted, search posts or hashtags instead and take the authors.
A creator search endpoint queries a platform's own user directory for a keyword and returns matching profiles: handle, display name, follower count, bio, verification and profile link. It is the discovery half of influencer research. Engagement rates, contact details and audience demographics are separate jobs on separate endpoints.
Not directly. These endpoints search by keyword; ask your agent for a larger result set and let it filter on the follower counts that come back.
No. Contact details are a separate job. Some creator bios carry a business email, and the people enrichment endpoints can resolve a work address from a name and company.
The one your audience uses. The comparison below shows who serves each platform and what a search costs there; treg.to does not pick a platform for you.
For most of these platforms it is a data provider reading the public index, not the platform's own API. Each row names the provider and links its documentation.