The cheapest social media data APIs, ranked

"Cheapest" is a harder question than it looks, because the advertised rate is only one of three numbers that decide what you pay. The other two are the floor you pay before your first call, and what happens to your bill when a call comes back empty. Rank on the advertised rate alone and you get one order. Rank on what a real workload actually costs and you get a noticeably different one.

Start here, because it is where everybody starts. Each figure is that provider's own published rate normalised to dollars per thousand results, with the date it was checked printed underneath.

A monthly minimum or a required subscription is money spent whether you call or not, so it belongs in the cost per result rather than in a footnote. Divide the floor by your monthly volume and add it to the rate. At a million results a month this changes almost nothing. At fifty thousand it can dominate: a floor in the hundreds of dollars against a small volume works out at several dollars per thousand on its own, before a single result is counted. This is why the ranking is different for different people, and why a single "cheapest" answer is usually wrong. The provider that is cheapest at scale is frequently among the most expensive at evaluation volume, which is exactly when people are choosing.

If you are billed per request rather than per result, divide by your fill rate. A list where 70% of targets resolve makes a per-request rate about 43% higher in real terms than it looks. Fill rate is workload-specific and it is not published by anybody, which is the honest reason this post cannot rank providers definitively. It is also the largest single variable — larger than the spread between most of the rates in the table above.

Some providers bill in credits that buy different amounts on different endpoints — a profile might cost one credit and a page of comments five. That is defensible pricing, and it makes the headline credit rate close to meaningless without knowing your endpoint mix. The way to handle it is to price your actual mix rather than the rate card: take the endpoints you will really call, in the proportions you will really call them, and work out the credits. It takes twenty minutes and it is the only version of this comparison that is about you.

For a team running high volume on paged endpoints with clean targets, the advertised rates are roughly the right ranking, and the providers with floors become competitive because the floor amortises away. For everybody else — evaluating, prototyping, running seasonal work, or pulling from lists with a meaningful dead fraction — providers with no floor and per-result billing win by more than the table suggests, and the gap widens the smaller and messier the workload is. We are in the first group's answer and we are firmly in the second's, which is worth saying plainly rather than claiming to win everywhere.

Write down your monthly volume, your endpoint mix and your best guess at fill rate. Add the floor. Then run a hundred real targets through the two or three providers that survive, and compare completeness as well as cost. A provider that is 20% cheaper and returns 15% less is not cheaper. That comparison takes an afternoon and no pricing page can do it for you, including ours.