Stop counting requests.
Start sending them.

Enterprise is 5,000,000 requests a month for $900, on infrastructure dedicated to your account. Unlimited removes the monthly cap entirely for $1,400. Both include dedicated infrastructure, unlimited API keys, no per-minute rate limit, priority support, and a Slack channel. If your volume is predictable and under 5M, Enterprise is the cheaper answer. If you are ingesting continuously and cannot put a ceiling on it, Unlimited is the one.

Fair use is 5–35 requests per second, sustained, 24/7 — no burst window, no daily cap, no monthly quota. At the top of that range you can pull roughly 90.7 million requests a month and the bill stays $1,400. We size your dedicated infrastructure around your expected pattern during onboarding, so the number you agree to is the number you get.

No. One request is one credit, on every endpoint and every platform, no matter how much data comes back or how hard it was to fetch. This is the thing worth checking when you compare headline prices: most scraping APIs bill in credits that are not 1:1, so a single call can cost 5, 10, 25 or even 75 credits depending on the target site, whether JavaScript rendering was needed, or which proxy pool it landed on. A plan that looks cheaper per credit can work out several times more expensive per actual request. With FetchLayer the number you divide by is always 1.

Because not every request costs the same to serve. A single post lookup is cheap and runs at the top of the range; a deep multi-page comment thread or a broad search across many communities is heavier and settles lower. We work out where your specific workload lands on the call, rather than quoting one number that would be wrong for half of it.

We talk before anything breaks — nobody gets a surprise cut-off. Once the 5,000,000 are used your account falls back to your credit balance, so requests keep working rather than failing. From there you have three ways to go: keep topping up at the standard rate of $1.99 per 1,000 requests, move to a custom plan sized to the volume you are actually running, or move to Unlimited if we still have capacity for it. If you are consistently over 5M, one of the last two is cheaper than topping up — we would rather put you on the right plan than bill you for overages.

Enterprise at $900 is a standing plan, not a promotion — it is our normal price for 5,000,000 requests and we are not planning to move it. Unlimited at $1,400 is the limited-time one: that rate is locked for three months in full, and nothing changes mid-period.

It depends on what you turn out to actually use — which is a more honest answer than a number we would have to walk back later. After three months we both know your real volume, and the renewal is priced against that rather than against a guess. If you genuinely are running unlimited, continuous ingestion near the top of the throughput range, the price goes up: reserving that much dedicated capacity indefinitely is not something we can promise at $1,400. If it turns out you need less than truly unlimited, which is the common case, it can go down, stay exactly the same, or rise slightly — it depends on where your volume settles, how heavy your calls are, and what infrastructure you are actually using. Either way you get your specific number in writing before the three months are up, and because the plan is month-to-month you can decline it and walk away owing nothing. Enterprise at $900 is unaffected, so dropping down to it is always an option.

Month-to-month, both tiers. Cancel whenever. The 3-month lock on Unlimited is a commitment from us to you, not a minimum term you owe.

Every platform FetchLayer supports, and every platform we add in future. Today that is Reddit, X/Twitter, App Store reviews, and YouTube comments, along with the MCP server for AI agents. New platforms land on your plan automatically as we ship them — same credentials, same bill, no upgrade and no per-platform pricing.

Your dedicated infrastructure is usually live within 1 to 3 hours of checkout. Nothing is blocked while we provision it — your API keys work straight away on our shared infrastructure, and we move you across once yours is ready. If you want an onboarding call, it happens the same day at whatever time suits you: throughput tuning for your workload, and a walkthrough of the endpoints relevant to your use case.

No. "Start instantly" takes you straight to checkout on either tier and your account is upgraded the moment it clears — no call, no quote, no procurement round-trip. We still recommend the 15-minute call if you want your volume and use case confirmed first, and onboarding happens either way.