Keyword research APIs you can pay per call

What keyword research APIs do

A keyword research API helps decide which search terms deserve a page, a campaign or a product name. The work has three parts. Expansion turns a seed term into candidate keywords. Measurement attaches numbers to those candidates, such as monthly search volume, cost per click, advertising competition or a ranking difficulty score. Trend analysis shows whether interest is rising, seasonal or concentrated in a region.

Those outputs answer different questions, and they come from different evidence. A volume figure from advertising data estimates how often a term is searched in a market. A Google Trends score is an index from 0 to 100 relative to the strongest point in the requested window, so two requests with different windows can give the same term different scores. A list produced by a language model from its training knowledge is neither: it can be a useful brainstorm, but it does not measure search demand.

An agent that plans content needs to keep that distinction in the data it stores. A planning table that mixes observed volume, a relative index and a generated guess in one column will rank terms by numbers that do not share a unit. Record where each value came from, the location and language it applies to, and the date it was retrieved.

Keyword research also borders other SEO tasks. A SERP check, a rank tracker, a backlink report and an on-page audit all start from keywords or domains, but they return positions, links and page diagnostics rather than demand for a term. The comparison below keeps those products out of the keyword quote even when the same provider sells both.

Compare the returned data, scope and limitations

Metrics routes return numbers for terms you already have: volume, CPC, competition, difficulty or intent. Ideas routes return new candidate terms from search data, such as autocomplete expansions, related queries or questions. Trends routes return relative interest over time or by region. A route can belong to more than one type when its description promises both, for example ideas delivered with volume and CPC.

Domain routes start from a website rather than a seed term. They list the keywords a domain ranks for, or the overlap between two domains, with positions and volume. They are useful for competitive research, but they describe where a site already appears rather than which new terms have demand, so they stay adjacent to the main comparison.

Modeled and generated lists stay separate as well. A deterministic query-variant generator and a model-knowledge estimate can suggest wording, yet they do not report observed search counts. A modeled demand indicator built from other signals is closer to data, but it still measures something other than search volume. A modeled or generated route cannot supply the row price of a keyword research service in this comparison.

CONNSKILL Growth Services name Google Ads data in matched endpoint descriptions. Google Trends SEO Keyword Data and CONNSKILL Growth Services name Google Trends. CONNSKILL Growth Services name DataForSEO signals for a modeled indicator. 2 describe location or country scoping in a matched route, and 1 describe suggestions ranked across Google, YouTube and Bing. A named source is the provider's statement, not an audit of freshness or method.

Each price is the cheapest endpoint of that service for that kind of data: the same number wherever this page quotes it.

Compare 3 keyword research services

Each row keeps the selected keyword endpoint and its reference price. The order follows observed paying wallets for the service; it does not rank the quality of the underlying data.

$/CALL is the price of the endpoint named under it: the cheapest keyword metrics: search volume, cpc, competition or difficulty, keyword ideas from search data: related terms, autocomplete or questions and relative search interest from google trends endpoint of the service or, for a service without one, its cheapest keyword research endpoint. Never the cheapest tool of a gateway. Other kinds of data have their own price in the tables below. KEYWORD RESEARCH ENDPOINTS counts the keyword research endpoints out of all the endpoints the service lists. BUYERS 30D counts distinct wallets that paid the whole service on-chain in 30 days, not only its keyword research endpoints. "Generalist" marks a gateway that sells more than three unrelated kinds of tools: it is never called the cheapest.

How to choose a keyword research endpoint

Start from the decision

A content team choosing between twenty topics needs volume and difficulty for those twenty terms. An agent exploring a new market needs expansion first and measurement second. A seasonal business needs a timeline. Write down which of those decisions the call supports before comparing prices, because the cheapest route may return a different kind of answer.

Check the unit of each number

Read whether a volume is monthly, averaged or a history, and which location and language it covers. Check whether competition describes advertisers bidding on a term or the difficulty of ranking organically. Check whether a Trends score is scaled to the request window. Keep the unit with the value so a later calculation does not compare a cost per click with a 0-100 index.

Keep the seed and the settings

Store the seed term, location, language, date and route with every result. Keyword lists change with each of those settings. Without them, a repeated request that returns different ideas looks like market movement when it is only a different query.

Plan the batch size

A route can price one seed per call, while another accepts a list of terms. A metrics call that accepts many keywords can cost more per call and less per keyword. Compare the cost of the whole list you need to size, including follow-up calls for ideas you decide to keep.

Test the operating limits

The monitored median uptime is 99.9% and the median selected quote is $0.03. Monitoring latency describes our checks, not a benchmark of a full keyword workflow. Try a representative seed, an empty or unusual term and a location you actually target, and note how the provider reports partial or missing data.

Choose by the output your workflow needs

Start with the response your next step can use. The candidates below use the same type classification as the comparison table. Each quote belongs to the lowest priced endpoint of the requested types, with the path kept.

Size demand for a keyword list

Send the exact terms, location and language you plan to target. Keep volume, CPC, competition and difficulty in separate fields with their source, because advertising competition and organic ranking difficulty measure different things.

Candidates today, most buyers first: CONNSKILL Growth Services ($0.03/call, 19 buyers) · Citable ($0.03/call, 2 buyers)

Expand a seed keyword

Treat the expansion as a candidate list. Deduplicate close variants, keep the seed that produced each idea and attach metrics before a term enters a content plan.

Candidates today, most buyers first: Google Trends SEO Keyword Data ($0.03/call, 42 buyers) · CONNSKILL Growth Services ($0.03/call, 19 buyers) · Citable ($0.005/call, 2 buyers)

Compare interest over time or by region

A Trends score is relative to the strongest point in the requested window or region. Store the window and geography with every value and compare terms only inside the same request.

Candidates today, most buyers first: Google Trends SEO Keyword Data ($0.03/call, 42 buyers) · CONNSKILL Growth Services ($0.03/call, 19 buyers)

See the keywords a domain already ranks for

A ranking snapshot belongs to a date, location and language. Save it with those fields so a later comparison measures movement rather than a change of settings.

Candidates today, most buyers first: CONNSKILL Growth Services ($0.05/call, 19 buyers) · Citable ($0.04/call, 2 buyers)

What the providers describe

These paragraphs attribute output contracts and limitations to each provider. Reference prices follow the matching endpoint type rather than an amount written in a marketing description.

Citable

Citable lists separate keyword routes. /v1/keyword-suggest describes autocomplete expansions, related queries and questions for a seed, ranked across Google, YouTube and Bing; its ideas quote starts at $0.005. /v1/keyword-metrics describes search volume, CPC, competition, difficulty and intent for up to 20 keywords per call, and the metrics quote starts at $0.03. A suggestion list and a metrics table are different purchases: the cheaper call names candidates, and the metrics call sizes them. The provider also states that a 4xx or 5xx response cancels the payment instead of settling it. That is the provider's policy statement; check the result record of your own calls rather than assuming every failure mode is covered.

5 keyword research endpoints of 17 · $0.005/call (GET /v1/keyword-suggest) · SOL · uptime 30d 100% · 199 ms · 2 buyers in 30 days

CONNSKILL Growth Services

CONNSKILL describes /v1/keyword-ideas as suggestions for a seed with search volume and CPC, at $0.03. Its /v1/keyword-metrics route checks up to 1,000 chosen terms with volume, CPC, advertising competition and monthly history from Google Ads data, and the endpoint description itself warns that advertising competition is not organic SEO difficulty. Difficulty is a separate route, as is a twelve-month Google Trends timeline quoted at $0.03. The provider also lists a modeled AI question-demand indicator based on DataForSEO signals and says it is not assistant query logs; keep that number apart from search volume.

9 keyword research endpoints of 41 · $0.03/call (POST /v1/keyword-ideas) · Base · uptime 30d 99.1% · 155 ms · 19 buyers in 30 days

Google Trends SEO Keyword Data

Google Trends SEO Keyword Data exposes four Trends views: interest over time, interest by region, related queries and related topics. Its relative-interest quote starts at $0.03, and related queries also count as keyword ideas at $0.03. The descriptions use 0-100 scores and percentage growth for rising queries. Those scores are relative to the highest value in the same request, not search counts, so they show direction and seasonality rather than how many people searched.

4 keyword research endpoints of 4 · $0.03/call (GET /related-queries) · ARB/Base/POL/SOL · uptime 30d 99.9% · 249 ms · 42 buyers in 30 days

Cost per call and for a larger keyword list

The median selected route costs $0.03 per call. At the same request shape, 100,000 calls total $3,000.00. Use the generated table as a reference calculation, then count how many calls one finished keyword plan really needs.

Each service at the price of its main keyword research endpoint, the one named in the table above.

A keyword plan usually takes more than one call. Expansion produces candidates, metrics size them, and a trend or domain check confirms the shortlist. Each step has its own route and price, and a route that accepts a list of terms changes the cost per keyword. Keep the exact endpoint, parameters and quote with each result, so a cheap suggestion call is not mistaken for the cost of a measured list.

Set a budget per plan and a stopping rule for repeated expansion. Separate an invalid term, an empty result and an unavailable upstream in the result record. The provider's terms decide whether those outcomes are charged. Read the payment challenge for the exact request before authorizing it, and do not estimate a keyword plan from the provider's cheapest unrelated SEO tool.

Inspect a request and its payment evidence

For keyword metrics, Citable documents GET /v1/keyword-metrics, with a metrics reference quote of $0.03. Read its current schema to pass the terms, location and language you need, and verify that the response carries the fields your plan uses.

The listed method and description establish the advertised contract. This explanation does not reproduce an HTTP 402 challenge or certify a paid response. Open the service page's payment evidence and obtain the challenge for the exact request before authorizing payment. Compare resource, method, input, network, asset and amount together; do not reuse a challenge from a neighboring audit or SERP tool.

How this comparison is built

The comparison starts with services that are online and payment-ready, have a known price, a description of at least 80 characters and measured 30-day uptime of at least 90%. Task selection then reads the documented endpoints. A keyword phrase in the service description can make a service a candidate without proving that a captured route returns keyword data. Pages below the index threshold can still be read directly.

Provider claims were reviewed against the public service and endpoint descriptions captured on October 22, 2026. They remain provider claims. Monitoring supplies operating figures and reference prices; it does not measure the accuracy of a volume estimate or the quality of a suggestion. Buyer counts describe observed paying wallets for the service, with the coverage limits explained on its linked page.

The endpoint classification supplies the comparison quote, per-type prices and job candidates. A row needs a fixed quote for keyword metrics, keyword ideas from search data or relative Trends interest. Domain snapshots and modeled or generated lists remain adjacent types and cannot qualify a service on their own. Keep the named path when comparing services: an adjacent route can have a different input, unit and evidentiary value on the same host.

Questions

How to pay

Each service answers an unpaid call with HTTP 402 and its price; your agent signs a USDC payment on the listed network and retries. No API key, no account, no subscription. See what x402 is.