Best Keyword Research API 2026: 9 Keyword Data APIs Compared

A keyword research API returns search volume, difficulty and related terms as JSON, so your code or AI agent can pull keyword data in bulk. Our recommendation for 2026: use DataForSEO for bulk keyword analysis, for feeding an AI agent, and as the cheapest paid option, at $0.09 for up to 1,000 keywords per live call. Use Ahrefs if you want the richest metrics, Semrush for the most countries, and the Google Ads API if it has to be free.

Below are seven keyword data APIs plus the two SERP APIs people often reach for instead, with entry prices, bulk limits, country coverage, the minimum spend to look up 10,000 keywords, and real data on how much search volume these APIs miss.

What Is the Best Keyword Data API in 2026?

There is no single best keyword data API. There is a best one for the job you are doing:

If you need	Use	Why
Bulk keyword analysis	DataForSEO	Up to 1,000 keywords per task, billed per request, no plan to buy
Keyword data for an AI agent	DataForSEO, or an MCP server built on it	Clean JSON, a cost field on every response, pay only for what the agent uses
The cheapest paid option	DataForSEO	$50 minimum top-up that never expires; 10,000 lookups cost about $0.90
The richest metrics	Ahrefs API v3	Traffic potential, parent topic and a backlink-based difficulty score
The most countries	Semrush API	143 country databases, per Collaborator's August 2026 API review
A cheap all-rounder with a platform behind it	SE Ranking	Pay-as-you-go API credits from $50
App store and marketplace keywords	Keywordtool.io	App Store, Play Store, Amazon and YouTube, not just Google
Free	Google Ads API	Keyword Planner data behind an API, no difficulty score
Live search results, not volume	SerpApi or Serper	The current Google results page as JSON

If you want keyword data inside an AI assistant like Claude Code or Cursor without signing up for an API and managing its key, Jack's SEO MCP gives your agent the same DataForSEO volume and difficulty as part of a paid plan.

Disclosure: Jack's SEO MCP buys the keyword data it gives your agent from DataForSEO, so we are a paying customer, not a partner. We chose it on price and data terms, and every DataForSEO figure here comes from its public pricing page.

How Do Keyword Research APIs Compare on Price?

Entry price here means the cheapest way to make your first API call, not the cheapest dashboard plan.

API	Entry price for API access	Billing model	Difficulty score?	Best for
DataForSEO	$50 top-up, then per request	Pay as you go	Yes (Labs API)	Builders, SaaS, AI agents
Ahrefs API v3	Lite plan, $129/month	Subscription + API units	Yes	Teams already on Ahrefs
Semrush API	Advanced plan, $549/month + unit packs	Subscription + units	Yes	Agencies, many markets
SE Ranking	$50 for 250K credits	Pay as you go or add-on	Yes	Budget all-rounder
Keywordtool.io	$88/month (Starter)	Subscription, 50 requests/day	No	Long-tail and autocomplete
Moz API	Free tier, paid from about $20/month (unverified)	Rows per month	Yes	Light use, Moz metrics
Google Ads API	Free	Developer token	No	Free Google volume
SerpApi	250 free searches, then $25/month	Searches per month	No volume at all	Real-time SERPs
Serper	2,500 free queries, then $50 for 50K	Prepaid credits	No volume at all	Cheap SERP checks

Sources: DataForSEO pricing, Ahrefs pricing and its API cost breakdown (rechecked 30 September 2026), the Semrush API cost breakdown, SE Ranking API, SerpApi pricing, Serper, and Collaborator's review for Keywordtool.io. Prices are in US dollars and move often, so check the provider's page before you commit.

What Does Each Keyword Data API Actually Give You?

DataForSEO

DataForSEO sells raw SEO data with no dashboard on top. Google Ads search volume costs $0.06 per task in the standard queue or $0.09 live, and one task holds up to 1,000 keywords. Its Labs endpoints add keyword difficulty, related keywords and search intent at $0.012 per task plus $0.00012 per result. The downside is the docs: there are hundreds of endpoints and the response shapes are verbose, so budget an afternoon of reading before your first useful call.

Ahrefs API v3

Ahrefs' help centre says API v3 works on Lite and above, and that extra units can only be bought on Enterprise. Lite is $129/month with 200,000 units, Standard $249 with 800,000, and Advanced $449 with 2,000,000, per the September 2026 breakdown above. Every request costs at least 50 units, and the same pool also pays for Ahrefs' MCP server. You get the metrics people pay Ahrefs for: keyword difficulty, traffic potential and parent topic. If you plan to build a product on the data, read Ahrefs' API terms first.

Semrush API

According to the Semrush API cost breakdown, which checked Semrush's own pages in August 2026, the pricing page lists API access on the Advanced plan at $549/month, while the developer docs still point to the legacy Business plan at $499.95. Either way, the plan comes with zero API units. You buy unit packs on top, starting at 2,000,000 units, and unused units expire at renewal. Semrush does not publish the pack prices, so you only see the real cost once you are logged in and paying for the plan.

SE Ranking

SE Ranking's pay-as-you-go API starts at $50 for 250,000 credits, or there is an API add-on from $149/month on top of a Core ($129) or Growth ($279) plan, per its API page. It returns volume, difficulty, keyword ideas and SERP data, and the free trial includes API access.

Keywordtool.io

Keywordtool.io builds long-tail lists from Google autocomplete, plus YouTube, Amazon, Bing, the App Store and Google Play. API plans start at $88/month. According to its API docs, the Starter plan allows 50 requests a day, shared with the web app and its MCP server, and a search volume request takes 1 to 1,000 keywords. Volume is exact for Google and Bing and estimated for the other platforms. It is the one to pick for bulk keyword research across iOS and Android apps, but it gives no difficulty score.

Moz API

Moz has a small free tier and paid plans that third-party pricing trackers put at about $20/month for 3,000 rows. We could not confirm those numbers on Moz's own site, so check its API page before you budget. Pick it if you want Moz's difficulty and Domain Authority numbers in the same call.

Google Ads API (Keyword Planner)

The Google Ads API exposes Keyword Planner data for free. You need a Google Ads account and a developer token with Basic access, which you apply for and which allows 15,000 operations a day. Per Google's historical metrics docs, you get average monthly searches, up to four years of monthly volumes, competition and bid ranges. There is no organic difficulty score, close variants get merged into one row, and Google refreshes the figures monthly.

SerpApi and Serper

Neither returns search volume. They return the live Google results page as JSON: titles, URLs, snippets, People Also Ask and SERP features. SerpApi has 250 free searches a month, then $25 for 1,000 or $75 for 5,000. Serper is cheaper per query at $50 for 50,000. Use one alongside a volume API when you need to see who ranks today.

What Does It Cost to Pull Volume for 10,000 Keywords?

Headline plan prices hide the real number, so here is one job priced on each API: search volume for 10,000 keywords, once a month. "Minimum spend" is what you must pay before the job can run.

API	Minimum spend	Metered cost of the job
DataForSEO	$50 top-up (credit never expires)	$0.90 live (10 tasks × $0.09), $0.60 standard
Google Ads API	$0	$0, within 15,000 operations a day
SE Ranking	$50 credit pack	Depends on credits per keyword, which the API page does not list
Keywordtool.io	$88/month	Fits only if your batches fit in 50 requests a day
Ahrefs	$129/month (Lite)	Fits in Lite's 200,000 units only if each keyword row costs 20 units or fewer
Semrush	$549/month + a unit pack	Unit pack prices are not published
SerpApi, Serper	Not applicable	Neither returns search volume

Only two APIs give you a hard number for this job: $0.90 and $0. SE Ranking needs a credit pack and the rest need a plan, and none of those four publishes enough to price the job exactly. Moz is left out because we could not verify its row prices. If your code or agent makes calls in bursts, per-request billing means you never pay for idle capacity.

Which API Is Best for Bulk Keyword Analysis?

For bulk keyword analysis, DataForSEO is the best fit: one task takes up to 1,000 keywords, and you pay per task rather than per seat. What limits a bulk job is usually the batch size or the daily cap, not the headline price:

API	Keywords per request	Daily or monthly cap
DataForSEO	Up to 1,000 per task	None beyond your credit balance
Keywordtool.io	1 to 1,000 per volume request	50 requests a day on Starter, 15 a minute
Google Ads API	Many keywords per request	15,000 operations a day on Basic access
Ahrefs API v3	Billed in units, at least 50 per request	Your plan's monthly unit pool
Semrush API	Billed in units	The unit pack you bought, which expires at renewal

A one-off list of 50,000 keywords is 50 DataForSEO tasks, about $4.50 live from the $50 minimum top-up, or exactly one day's Starter allowance on Keywordtool.io. For a monthly refresh of the same list, cache the results and only re-query terms you have not seen, since Keyword Planner figures change monthly anyway.

Which Keyword APIs Cover the US, Europe and Other Countries?

All nine APIs here are global in the sense that you pass a location with each request, and the US is the default in most examples (DataForSEO's location_code 2840, in the request below, is the United States). Search volume is always per country and language, so "Europe" is not one market. To size a keyword across Europe, query the countries you care about one by one.

• Google Ads API: takes up to 10 geo targets per request, per Google's historical metrics docs, so you can combine several countries into one figure.
• Semrush API: the widest database list at 143 countries.
• DataForSEO, SE Ranking, Ahrefs: country is a request parameter. Check each provider's location list for smaller markets before you build on it.

If a market is missing or its numbers come back null, that is a coverage gap, not proof of zero demand (more on that below).

How Accurate Is the Search Volume an API Returns?

Less than the clean JSON suggests. We ran this page's target keyword and the four related queries that bring it the most impressions through Keyword Planner data (via DataForSEO, US, 19 October 2026), then compared them with what Google Search Console recorded for this one page over the previous 28 days.

Query	Monthly volume from the API	Impressions to this page (28 days)
keyword research api	50	None recorded
api keyword research	No data	148
best keyword data api	No data	117
keyword research tools api	No data	86
best keyword research tool api	No data	86

Four of the five came back as null: below Keyword Planner's reporting floor. Yet those four sent 437 impressions to one page at average positions between 19 and 63, while the one query with a measured volume recorded none. A "no data" result is not zero demand. It means the source cannot measure it. For your own site, the free Search Console API shows real queries that no keyword API will.

What Does a Keyword Research API Request Look Like?

Most of these APIs are a single HTTP call. Here is a search volume lookup against DataForSEO:

curl -X POST "https://api.dataforseo.com/v3/keywords_data/google_ads/search_volume/live" \
  -u "$DATAFORSEO_LOGIN:$DATAFORSEO_PASSWORD" \
  -H "Content-Type: application/json" \
  -d '[{"keywords": ["keyword research api", "api keyword research"], "location_code": 2840}]'

The response, trimmed to the fields that matter. The volumes are the ones we got on 19 October 2026, and cost is the live price DataForSEO listed that day:

{
  "tasks": [{
    "cost": 0.09,
    "result": [
      { "keyword": "keyword research api", "search_volume": 50 },
      { "keyword": "api keyword research", "search_volume": null }
    ]
  }]
}

Two things to build for. First, search_volume can be null, so never coerce it to 0 or your scoring will bury real long-tail terms. Second, the cost field reports the price of every call, so log it per run and you will catch a runaway loop before it costs you.

How Do You Choose a Keyword Data API?

Work through these in order and stop at the first one that rules a provider out:

1. Do you need difficulty, or only volume? Volume only means the Google Ads API (free) or DataForSEO's Google Ads endpoint. Difficulty means DataForSEO Labs, SE Ranking, Ahrefs, Semrush or Moz.
2. Are your calls bursty or steady? Bursty and automated favours pay as you go (DataForSEO, SE Ranking). Steady, heavy use can justify a plan.
3. Will you store or resell the data? If you are shipping a product, read the provider's API terms before writing code.
4. Do you already pay for Ahrefs or Semrush? If yes, check your plan's API allowance first. You may already own enough units.

Comparing dashboards instead? Our roundup of the best keyword research tools covers that side, and the Keyword Surfer alternative guide covers free volume sources.

How Do You Use a Keyword Research API From Your Repo With an AI Agent?

The API call is the easy half. Getting the numbers into the agent that writes your posts is the useful half. This loop works with any provider above:

1. Put the API login in an environment variable, not in a file your agent can commit.
2. Ask the agent to look up your seed keyword and its close variants for your target country, and save volume and difficulty to a JSON file in the repo. That file is your cache.
3. Have it pull the live top 10 Google results for the best candidate.
4. Give it a brief: search intent, headings, where the keyword goes, which sources to cite. It writes content/blog/<slug>.md with frontmatter at the top.
5. Check the file before you commit: title and meta length, one H1, heading order, links that resolve. A script that exits non-zero is enough.

That is an evening of wiring, plus upkeep. If you'd rather not build it, Jack's SEO MCP is the same loop already built, running from inside your repo. Connect it to Claude Code, Cursor or Codex and your agent gets live keyword volume and difficulty, a brief and 60+ checks built on Google's published guidance, with managed keyword data on every paid plan. It writes each article as a markdown file straight into your content folder. Our guides on keyword clustering and keyword strategy cover what to do with the data.

What Are Common Mistakes When Picking a Keyword Data API?

• Buying a plan to test an idea. Semrush and Ahrefs bill monthly before you have made one call. Prototype on a pay-as-you-go API, then upgrade if the richer data earns its keep.
• Trusting one source for volume. Volume is an estimate, and providers model it differently. Treat it as a range.
• No cache. Keyword Planner figures only change monthly. Cache results so you never pay twice for the same lookup.
• No spend logging. An agent stuck in a retry loop can burn credit fast. Log the cost of every call.

Frequently Asked Questions

What is a keyword research API?

A keyword research API is a web service that returns keyword data as JSON instead of showing it in a dashboard. You send one or more keywords and a target country, and it returns metrics like monthly search volume, keyword difficulty, cost per click and related terms. Developers use one to automate research, build SEO tools, or feed keyword data to an AI agent that writes content.

Which keyword research API should I use?

Use DataForSEO unless you have a reason not to. It is the best keyword data API for most developers in 2026: it bills per request from a $50 minimum top-up, and a live volume lookup for 10,000 keywords uses about $0.90 of that credit. Pick the Google Ads API if it must be free, Ahrefs if you want the richest metrics, and Semrush if you need the most countries.

Which keyword API should an SEO professional use?

An SEO professional who already pays for Ahrefs or Semrush should check their plan's API allowance first, since Ahrefs includes API units from Lite upwards. If you do not, or you need bulk volume and difficulty for client work without another monthly plan, DataForSEO is cheaper per keyword and has no seat to buy.

What is the best API for bulk keyword analysis?

DataForSEO is the best API for bulk keyword analysis for most people: each task holds up to 1,000 keywords and there is no daily cap beyond your credit. Keywordtool.io also takes 1,000 keywords per volume request but caps Starter at 50 requests a day. The Google Ads API is free but limited to 15,000 operations a day on Basic access.

Which keyword data API works best with an AI assistant?

Any keyword data API that returns clean JSON works with an AI assistant, and DataForSEO is the usual choice because it bills per call and reports the cost of each one. If you would rather not manage an API key at all, an MCP server such as Jack's SEO MCP gives Claude Code, Cursor or Codex keyword volume and difficulty directly.

Do keyword research APIs cover countries outside the US?

Yes. Every keyword research API here takes a location with each request, so you can pull volume for the UK, Germany or any other supported country, not just the US. Semrush lists the most country databases at 143. Volume is per country, so for a Europe-wide figure you query each country, or combine up to 10 geo targets in one Google Ads API request.

Is there a free keyword research API?

The Google Ads API is the main free keyword research API. It returns Keyword Planner data, including average monthly searches and up to four years of monthly volumes, at no cost. Access needs a Google Ads account and a developer token approved for Basic access. The Google Ads API has no organic difficulty score, so pair it with a paid source if difficulty matters to you.

Does the Ahrefs API cost extra?

The Ahrefs API costs nothing on top of a Lite plan or higher, but it is not available on Starter. Ahrefs API v3 draws on a monthly pool of API units that comes with the plan (200,000 on Lite), every request costs at least 50 units, and extra units can only be bought on Enterprise. Price your real call volume against that pool before relying on the Ahrefs API.

Why does a keyword API return null search volume?

A keyword API returns null search volume when the term falls below Google Keyword Planner's reporting floor. That is common for long-tail and technical queries. Null means the provider cannot measure the demand, not that nobody searches the term. Store null separately from zero, and use Google Search Console data to see the real queries reaching your own pages.