Same job, different price: what data costs per provider

In my experience, most teams pick a search API or an email verifier once. They wire it in, the invoice arrives, and nobody looks at the unit price again. Meanwhile the same call is on sale somewhere else for a fraction of it.

I wanted to know how large that gap is, so I read the list price of five common data jobs across every provider in one catalog, on one day. The short answer is that the gap between the cheapest and the dearest provider for the same job was never below 25 times.

Before the numbers, the disclosure. I founded looot, and looot sells access to every provider in these tables. That is why I can read all of their prices in one place, and it is also a reason to check my work. Everything below comes from two public reads that need no account, so you can repeat it.

How were the prices read?

On 2026-10-11 I read the looot public catalog, the same list an agent searches before it runs a call. Each endpoint in it carries a job, a provider, a list price and a billing unit. A script took every endpoint for a job, kept the cheapest endpoint of each provider, and compared providers.

Three rules keep the comparison honest:

• Only endpoints billed per call are compared. An endpoint billed per result is listed with its unit and left out of the spread, because a call and a result are not the same unit.
• One row per provider. A provider with six endpoints for one job counts once, at its cheapest.
• The spread is the dearest provider divided by the cheapest. Nothing else is weighted in.

I made no paid call for this article. These are list prices, not measured bills.

How far apart are prices for the same data?

Far. The table has one row per job, with the number of providers that price it per call, the cheapest, the median and the dearest.

Two things stand out. Every job has a spread of 25 times or more. And in every row the median is below the halfway point between the cheapest and the dearest, so a few expensive endpoints stretch the range.

What does the same Google search cost at each provider?

This is the cleanest comparison of the five, because every endpoint does the same thing: you send a keyword, you get the organic results back.

Seven providers, and the dearest is listed at 27.8 times the cheapest. The two cheapest are almost level with each other. Six of the seven sit within about 4.4 times of the cheapest, and one sits far above the rest.

A higher list price does not mean a worse product, and a lower one does not mean a better one. The catalog descriptions show real differences in what comes back. One endpoint returns organic results only. Another adds the knowledge graph, the answer box and related questions in the same call. This table measures price and nothing else.

One caveat on this job. The catalog overview counted 11 of the 12 Google search endpoints as runnable on the day I read it, and the public read does not say which one was off. The table lists all seven providers at list price.

An earlier note in my own research quoted the same two ends of this range. I did not reuse it. Both numbers were read again on 2026-10-11 and both held.

What does email verification cost across providers?

Verification is also close to like for like. You send one address and get a verdict back.

Thirteen providers price it per call. Twelve of them fit inside a range of about 14 times. The thirteenth is listed far above the others and stretches the full spread to 99 times. I do not know why that endpoint is priced where it is, and I did not test whether it returns more.

The billing rules differ as much as the prices do. One provider's description says it charges nothing for a catch-all or unknown verdict. Another says it uses a credit on every attempt. Most descriptions state no rule at all. On a list with many catch-all domains that rule can matter more than the unit price.

Why are email finding and company enrichment harder to compare?

Because the endpoints under one job do not take the same input or return the same record. The tables still show the spread, but read the endpoint names before you draw a conclusion.

The cheapest email endpoint takes a LinkedIn profile, not a name and a domain. The dearest one unlocks a verified address for a person you have already identified. Those are different products that share a job label. Among the ten finders whose description names a name and a domain as the input, the range is much narrower, though one more endpoint with that input is priced far above them. I broke that out in a separate post on finding emails by name.

Misses are billed differently too. Two finder descriptions state that a search that finds nothing costs the same as one that finds an address. Four others state that a miss costs nothing. A provider with a higher unit price and free misses can be the cheaper one per found email.

Company enrichment has the widest spread of the five, at just over 100 times. The cheapest endpoint returns the company behind an email address. The dearest returns a deduplicated company record for a website. Four more providers bill per result and are listed at the bottom, outside the spread. The live B2B data API ranking orders the per-call providers by today's price.

What does fetching a web page cost?

Turning a URL into markdown is the cheapest job here in absolute terms, and it still has a spread of 25 times.

The dearest endpoint loads the page in a browser, which costs more to run than a plain fetch. That is one case where the catalog description explains part of the gap.

What are the prices right now?

The tables above are frozen on the day I read them. The block below is read from the catalog when this page loads, so you can see whether the cheapest price has moved since.

For the full list of endpoints behind any row, open the public catalog or the live SERP API ranking.

Why is the gap this large?

I can read prices. I cannot read reasons, so treat this section as my opinion.

• Providers do different amounts of work. A plain fetch and a browser session are not the same cost to run.
• Providers return different amounts of data under the same job name.
• Some list prices assume a volume plan on the provider's own site and look high when sold one call at a time.
• Few buyers compare. A price that is rarely checked has little reason to move.

The last one is the reason I think this data is worth publishing.

What does this not show?

• No success rate. The catalog publishes no measured success rate for these endpoints, so I quote none.
• No speed, no freshness, no field coverage.
• No vendor-site prices. These are looot list prices. A provider may charge more or less on its own site, and volume plans change the picture.
• No measured bill. Nothing was run.

A fair next step is a paid test: the same 100 inputs through every provider for one job, with hit rate and cost per hit. That has not been done.

How do I check a price before I pay?

Ask your agent to read the catalog and quote before it runs anything. The prompt below does that for one job.

Last checked 2026-10-11 against the looot public catalog. Written by Walid Boulanouar, who founded looot. The method is described in how to read a SERP price from the catalog and what 1,000 lookups cost.