FanCentro Profile Scraper

Extract public FanCentro creator profile data at scale — no login required.

What you get per creator

Field	Description
username, profileUrl	Public handle and profile link
displayName	The creator's display name
creatorId	Stable numeric ID — safe to key on when a handle changes
bio	Profile text
imageLinksExpireAt	ISO-8601 deadline for avatar and cover — the earlier of the two. null if the URL format changed and we could not read one. This is the exact per-creator figure; the ~52-59 minute range below is the population, not your row.
avatar, cover	Image URLs — see the note below, these expire
followers, subscribers	Audience counts
postsCount, videosCount	Content counts
hasSubscriptions	Whether the creator offers a paid subscription
perks	What a subscription includes: exclusiveContent, directMessages, clipDiscount
scrapedAt	ISO-8601 timestamp of when this row was read

What varies and what does not, measured across 120 distinct live creators so you can plan a segment before paying for one: exclusiveContent split 85/35, clipDiscount 36/84, and directMessages was true for all 120 — real, but useless to filter on. cover was present for 50 of 120; bio and avatar for all of them.

📏 bio is short — plan for that before you build on it. Measured on the same 120 creators:

bio length	share
median	35 words
≥ 25 words	75%
≥ 50 words	23%
≥ 120 words	3%

We return the bio exactly as FanCentro publishes it and never pad it. If your use case needs long-form text — indexable landing pages, content generation, semantic search — most profiles will not carry enough on their own. That is a property of the source, not of this Actor, and you should know it before you pay per profile rather than after.

🤖 About 1 in 4 bios is FanCentro's own template, not the creator's writing. Measured on 120 live creators, 2026-07-30: 31 of 120 (26%) open with stock phrasing — "Welcome to my VIP space! I'm …", "✨ This is my only official Fancentro", "Step inside to unlock my premium feed". They are genuine profile text and we return them verbatim, but they are auto-generated around the creator's name and post count rather than written by them.

It matters if you are doing anything language-shaped — tone analysis, clustering, embeddings, "find creators who describe themselves as X" — because a quarter of your corpus will cluster on FanCentro's copywriting instead of on the creators. Filter on those phrases if that is your use case. We deliberately do not ship a bioIsTemplate flag: the phrase list would go stale silently as FanCentro adds variants, and a flag that quietly stops being true is worse than one you maintain yourself against text you can see.

⏱️ avatar and cover are short-lived links — treat them as valid for well under an hour. FanCentro serves images from a CDN behind a signed URL with a built-in expiry. The remaining lifetime at the moment you receive a row varies: it is set when FanCentro renders the page, not when we read it, so a cached page yields a shorter window. Observed in practice: roughly 52 to 59 minutes from scrapedAt, and it can be less. After expiry the link returns HTTP 410 Gone — not 403. Measured 2026-08-06 on a link whose deadline had passed 16 hours earlier: HEAD and ranged GET both return 410. A 403 is a different answer: it is what the CDN says when the token is malformed, verified by tampering with the epoch and with the signature separately on a still-valid URL. So 410 = expired, 403 = rejected signature, and an ingest pipeline written to "treat 403 as expired" will misfile every genuinely expired link. The signature cannot be stripped or extended in either case.

⚠️ Re-running within ~10 minutes does NOT get you a fresh link — after that it usually does. Measured across two days:

gap between scrapes	what you get
45 s – ~8 min	the identical signature and deadline — the render is cached upstream
~10 min and beyond	a new signature with a full ~60 minutes of life

So the honest rule is narrower than "re-scraping is not a workaround", which is what this section said until 2026-08-06: a retry inside the cache window buys you nothing, a retry outside it does.

Either way, mirror the image at ingest. Re-scraping to refresh a link costs another paid row per creator, and the cache window means it may silently do nothing. Storing the URL and fetching it hours later is the one usage this field cannot support.

How much time you actually have, measured on a 174-profile run:

	remaining life when you read the dataset
first row of the run	51 min
last row of the run	59 min
shortest of all 174	49 min

The first row is the one that matters — it was fetched earliest, and the deadline is absolute, so a long run eats its own margin. At the observed 0.43 s/profile a 500-row run takes 3.6 minutes and its first row still has ~45 minutes left. That is why maxResults is capped at 500: a single run cannot get long enough to outlive its own image links. Walking the whole catalogue uses discoverOffset across several short runs instead of one long one — which keeps every batch's links alive by construction rather than by luck.

If you need the images, download them as part of the same run. Do not store these URLs and fetch them later. We would rather state this plainly than sell you links that quietly break.

Not included, because FanCentro does not publish it: subscription price is only visible behind a login, so it is not in this dataset. Neither is a photo count or a verified flag — FanCentro's public profile payload has no such fields. We would rather omit a field than ship one that is always the same value.

Input

{
  "usernames": ["nicolkremers", "xev-bellringer"],
  "discover": true,
  "maxResults": 10,
  "proxy": ""
}

• usernames — FanCentro handles (the part after fancentro.com/).
• discover — also pull creators from FanCentro's public creator listing, paging through it until your maxResults is met. Measured 2026-08-05: the catalogue is 8,185 creators, reached in 17 runs of 500 by following nextOffset. Two independent full walks returned the identical set — so the sweep neither skips nor repeats anyone, and you do not need to dedupe. On by default, so a run with no input at all still returns creators. ⚠️ Every discovered creator is a billable result, up to your maxResults.
• discoverOffset — where in the listing discovery starts. This is what lets a second run return different creators than the first. Leave at 0 for the most trending creators; to sweep the catalogue, set it to the nextOffset the previous run reported in its log.
• maxResults — hard cap on profiles returned, and therefore on what the run costs. Defaults to 10 so a first run is cheap; the maximum is 500. At $0.025 per profile, the default run costs $0.25.
• proxy — optional; your own proxy URL (a public HTTP/HTTPS proxy). Leave empty to use the default.

Output

One dataset item per creator that was successfully read. Use it for market research, creator discovery, and lead generation.

Handles that could not be read are not silently dropped: the run's status message and the RUN_SUMMARY record report how many you requested, how many were delivered, how many were not found or unreachable, and how many were skipped by the maxResults cap — so you can tell "this creator does not exist" apart from "we capped your run".

If a run returns nothing at all, it fails rather than reporting success with an empty dataset, and nothing is charged.

Pricing

Pay per result: $0.025 per profile scraped. You only pay for profiles actually delivered to your dataset — failed handles and capped handles cost nothing.

Notes

Public data only. Read-only. No login, no messaging, no account actions.

Privacy

This dataset describes real people, so it is personal data under the GDPR and you become an independent controller of any copy you keep.

• What it is — public profile information that FanCentro shows to logged-out visitors. Nothing behind a login or a paywall, and no media files are stored by us.
• Retention — every row carries scrapedAt. Please apply a retention limit rather than keeping profiles indefinitely, and re-run instead of relying on stale copies.
• Removal requests — a creator who wants to be excluded can write to the address on our Apify developer profile, or open an issue on this Actor. We maintain a denylist: once a handle is on it, this Actor stops returning that creator on any future run, for any buyer. Please honour such a request in copies you already hold.

---