Questions people actually ask

No. You never apply to a platform, wait for app review or maintain a developer account. Create a FastFetchz account, get one ff_live_ key, and the same key works for every platform in the catalog. We handle the upstream access; you only ever talk to fastfetchz.com.

Official APIs each need their own application, approval process, token lifecycle and response format, and several no longer offer public-data access to new developers at all. FastFetchz gives you one key, one consistent schema and one prepaid balance across all platforms, and every call is priced up front. The trade-off is honest: we return publicly visible data only, so anything the official API exposes about your own logged-in account is out of scope.

You can get anything publicly visible without logging in: profiles, posts, comments, videos, hashtags, search results, creator storefronts and shop product listings with their reviews, ad libraries and search-engine results. Follower and following lists are available where the platform exposes them — on X today. You cannot get private accounts, login-gated content, direct messages, email addresses or phone numbers, or anything about a user's own account. Coverage is per endpoint: the API reference lists exactly which fields each service returns and its measured page size.

The live list is on the API reference and in the Platforms menu — it is read from the catalog, so it is always current. Today it spans social networks, short video, creator link pages and storefronts, music, podcasts, search results and ad libraries, plus three publishing services (Bluesky, Telegram and Discord) that post on your behalf.

Every new account gets 100 free credits on signup, recorded in your ledger, which is enough to test any endpoint in the catalog. Sandbox responses are free and unlimited. No card is needed to start.

Profile, post and product endpoints are fetched at request time, so results reflect the platform now. On some endpoints the upstream fetch is allowed a short cache window to keep latency low; where we can tell honestly, the response carries a cached field of true or false. On those endpoints, adding ?fresh=true drops the cache window and forces a live fetch; on endpoints that are already fetched live it changes nothing. Either way it costs exactly the same as a normal call, and the default is false.

No. We store request metadata so you can audit usage — timestamp, key prefix, service, the target you asked for, status, latency and credits charged — but response payloads are not written to our database. The only exceptions are short-lived caches of platform identifiers (for example a username-to-ID mapping) that make repeat lookups faster. Your Playground runs display results in your browser at the time of the call.

A private, restricted or deleted target returns a clear error — 403 resource_restricted or 404 not_found — with zero credits charged. We do not attempt to bypass privacy settings, and we never log in to reach content. A public profile that simply has nothing to return (no posts yet, for example) is also not charged.

Partly, and each endpoint says which it is. The response envelope — success, service, credits_charged, credits_remaining, data — is identical on every service. Inside `data`, a defined set of endpoints is mapped to a shared schema: eight profile endpoints (Instagram, TikTok, X, YouTube, Facebook, Threads, LinkedIn profile and LinkedIn company) return one flat cross-platform profile record with exact counts rather than rounded display numbers, and a few more return a trimmed version of the upstream shape. Every other endpoint returns the upstream response as-is, and its reference page states that in the description. We would rather tell you the shape you will actually get than claim a uniformity that is not there yet. The Playground's Preview view only renders fields actually present in the response — it never guesses.

Yes. FastFetchz runs a standard remote MCP server with OAuth at https://fastfetchz.com/mcp. Add that URL as a connector in your assistant, sign in once in the browser, and the agent gets five tools: list platforms, list services, estimate cost, fetch, and check balance. We have tested the full flow with Claude, ChatGPT, Grok and Cursor; VS Code, Zed, Gemini CLI, Windsurf and Cline document remote MCP support and connect the same way. Step-by-step instructions per client are on the MCP server page.

Two mechanisms, both enforced on every request. First, the prepaid balance is a hard ceiling: when it hits zero the call returns 402 and nothing is billed later. Second, each API key can carry its own monthly call cap; once a key reaches it, further calls return 429 monthly_cap_reached until the next calendar month. Give each agent its own key with its own cap, and revoke any key instantly without affecting the others.

Yes, through their standard HTTP request modules — there is no dedicated FastFetchz node or app yet. Every endpoint is a plain HTTPS GET or POST to https://fastfetchz.com/v1/{platform}/{service} with your key in an X-API-Key header and a JSON response, which all three tools handle natively. Each service page shows a ready-made request you can paste in.

Not yet. FastFetchz is read-only today: every service fetches public data. Publishing to Bluesky, Telegram and Discord has been built but is withdrawn from sale until we have proven a real message end to end — it will appear in the catalog the moment it does.

You buy credits up front and one credit is worth $0.0015. Every successful call deducts that service's credits-per-call cost, which is shown on its page along with the equivalent price per 1,000 calls. Most endpoints cost 1 credit; a couple cost less and a handful cost more. There are no subscriptions, no platform fee and no monthly minimum.

No. Credits do not expire and there is no monthly reset. They are forfeited in only two cases: if you close your account, any balance beyond the 30-day refund window lapses; or if we terminate the account for breach of the Terms of Service or the Acceptable Use Policy.

Yes, on unused credits. Credits bought within the last 30 days can be refunded to the original payment method on request, minus any credits already consumed. Credits already spent on successful calls are non-refundable. Failed calls are never charged in the first place, so they never need refunding. Full terms are in the Refund Policy.

Never. 400 bad requests, 401 authentication errors, 402 insufficient-credit responses, 403 restricted targets, 404 not-found, 429 cap errors and 500 upstream failures all return without deducting credits. You only pay when we hand back data for the target you asked for.

From Billing in your dashboard: pick a credit pack and pay by card through our checkout. Credits land on your balance as soon as the payment confirms, with a receipt by email and an entry in your ledger. Larger packs carry a lower per-credit rate. If you need more volume than the published packs, contact us for a quote.

Three, all enforced on every request. Your prepaid balance is the hard ceiling. Each key can carry a monthly call cap you set yourself. And every account has a requests-per-second limit: 3 per second before your first purchase, then the figure printed on the highest pack you have bought (3, 10, 20 or 40). Exceeding it returns 429 rate_limited with a Retry-After header, and like every failed call it is never charged. The limit applies across REST, the Playground and MCP together.

As many as you need. A key per project, client or agent is a good habit — each one carries its own optional monthly cap and can be revoked instantly without affecting the others. Usage in your logs is broken down by key prefix.

No. Keys are hashed before storage and shown only once at creation, so we cannot recover one. Revoke the old key and create a replacement; your balance is on the account, not the key, so nothing is lost.

Sandbox responses are deterministic sample data shaped like live payloads, clearly marked as samples, so you can build and test your parsing for free. Live mode fetches real records and is the only mode that charges credits. Every response tells you which mode it was.

You are never charged for a failed call — that is the guarantee, and it holds without exception. Each service currently runs on a single upstream source, so if that source has an outage, calls to that service return an error and zero credits are deducted. Retry later. The Status page publishes measured success rates and response times per endpoint group from real customer traffic over the last hour, day and week. We do not have automatic failover to a second source today; when a service gains one, it will be stated here.

Collecting publicly available data is lawful in most jurisdictions, but how you use it is your responsibility. Our Acceptable Use Policy requires public data only, no login-gated content, no building of private-individual profiles for surveillance or harassment, and compliance with the data-protection law that applies to you — GDPR if you serve UK or EU residents. We are not able to give legal advice for your specific use case.

FastFetchz is a trading name of Nuvexa Media FZ LLC, a company registered in the Ras Al Khaimah Economic Zone, UAE, licence 17005908. Account data is processed under UAE Federal Decree-Law No. 45 of 2021 and, for UK and EEA customers, UK GDPR and EU GDPR. The Privacy Policy sets out the detail.

Use the Contact page or email support@fastfetchz.com. Include your key prefix and the timestamp of the request and we can trace it in the logs directly. We answer within one business day, usually much sooner.