By the end of this guide you will have a prioritized list of keywords for a site or a content plan, grouped by topic and matched to the pages that should target them. The job is done by a mix of AI keyword and SEO tools, which suggest and cluster terms, and search-results tools, which show you what is actually ranking.
If you want to do keyword research with AI, the key is to treat the AI as a fast source of ideas and groupings, then check every promising term against live search results yourself. This guide walks through the steps and compares seven tools from our Marketing and SEO category.
Key takeaways
- Start from a topic or competitor URL, expand into related terms, check what ranks for each, group them into clusters, then assign each cluster to a page.
- The main decision is the kind of tool: content optimization platforms (Clearscope, Search Atlas), AI agents that plan topics (Serplux, Lunroo, Creaitor.ai), or search-results APIs for developers (SerpApi, Serper).
- AI suggestions are not data. Check intent against the live results, and use a volume source you trust, since several of these listings do not describe volume figures.
- Free options are limited: SerpApi lists 250 searches a month, and Clearscope offers a 14 day free trial but no free plan.
What you need before you start
- A site or a topic. A URL, a product, or a subject area you want to rank for.
- A short goal. What you want visitors to do, such as read, sign up or buy.
- Competitor URLs. A few sites that already rank for your topic.
- A spreadsheet. A simple place to collect and sort keywords.
- Access. Check how each tool onboards new users; Lunroo, for example, has paused self-serve sign-up and onboards accounts by email. The API tools also need some comfort with code.
- Time. A first pass takes under an hour. A full plan for a site takes longer.
- Difficulty. Easy for the web tools, moderate for the APIs.
Which kind of AI tool fits this job
Choose by whether you want ideas, optimization guidance or raw search data.
Content optimization platforms. Clearscope has Topic Exploration, which generates keyword research from a topic or URL, and Drafts that analyze current search results and competitors to suggest terms and search-intent guidance. Search Atlas combines keyword tracking, site audits and content tools in one account, with AI visibility tracking on higher plans.
AI agents that plan and suggest. Serplux uses agents for strategy, competitor monitoring and weekly blog topic ideas. Lunroo scans a niche, checks the search results and picks keyword topics, then builds briefs inside a content plan. Creaitor.ai lists keyword research and AI keyword clustering alongside content writing and AI-assistant visibility tracking.
Search-results APIs for developers. SerpApi returns Google and other search engine results as structured data, including organic listings and Maps, Local and Shopping blocks. Serper is a Google Search API that returns organic results, People Also Ask questions and related searches as JSON. Use these when you want to build your own rank tracker or pull SERP data into a spreadsheet.
Step 1: Define your seed topics and goal
Good research starts with a narrow brief.
- Write down three to five core topics your site covers.
- Note who the audience is and what they want to do.
- List one or two competitor sites.
- Decide what counts as a good keyword for you: relevance to what you offer matters more than raw size.
Check it worked by being able to describe each seed topic in one line.
Step 2: Expand your seeds into candidate keywords
Let AI widen the list, and keep it in a spreadsheet.
- Enter a topic or a URL into a research feature. Clearscope's Topic Exploration generates keyword research from either.
- Ask an agent for ideas. Serplux suggests 50 or more blog topics every week based on trending searches, and Lunroo picks keyword topics after scanning your niche.
- Pull questions people ask. Serper returns People Also Ask questions and related searches, which are natural long-tail candidates.
- Collect everything in one sheet, and remove duplicates and terms that do not fit your business.
Check it worked by having a list of at least a few dozen candidates tied back to your seed topics.
Step 3: Look at what actually ranks
AI suggestions are only guesses until you see the live results.
- Search each promising keyword and look at the first page.
- Note the kinds of pages that appear: guides, product pages, lists, videos or forums.
- For scale or automation, pull results through an API. SerpApi returns structured results from Google and other engines, and Serper says it returns Google results as JSON in about one to two seconds.
- Record the result type next to each keyword.
Check it worked by being able to say, for each keyword, what kind of page ranks today.
Step 4: Judge intent, difficulty and value
Decide which keywords deserve a page.
- Match intent to your page. If the results are product pages and you plan a guide, skip or reframe the keyword.
- Look at who ranks. A page full of large, well-known sites is harder to beat than one with small sites.
- Use a volume source you trust. The listings for these tools do not all describe volume figures, so confirm what each one provides before you rely on it.
- Score each keyword by relevance to your goals, intent fit and how realistic ranking looks.
Check it worked by having a short priority list, with a reason beside each keyword.
Step 5: Cluster keywords and map them to pages
One page can target a group of related terms, so group them first.
- Group keywords that share intent and could be answered by one page. Creaitor.ai lists AI keyword clustering, though for a small list a spreadsheet works fine.
- Name each cluster by its main keyword.
- Assign each cluster to an existing page, or to a new page you plan to write.
- Keep clusters separate when the intent differs, so pages do not compete.
Check it worked by confirming that every priority keyword belongs to exactly one page.
Step 6: Build briefs from the research
Turn keyword clusters into content plans.
- For each page, write a short brief with the main keyword, supporting terms, intent and the questions to answer.
- Use the tool's analysis. Clearscope Drafts suggest terms and search-intent guidance based on the current results.
- Lunroo turns each topic into a brief inside a content plan with the angle, structure and internal links mapped out.
- Add what only you can say: your data, examples and product knowledge.
Check it worked by handing a brief to a writer and seeing whether they could start without asking questions.
Step 7: Track results and refresh the list
Keyword research is a loop, not a one-off.
- Track positions for the keywords you targeted. Serplux refreshes Google positions every 24 hours, and Search Atlas lists keyword tracking and site audits.
- Watch published pages over time. Clearscope's Content Inventory tracks how published pages perform.
- Consider AI answer visibility. Clearscope's Query Tracking watches queries across ChatGPT and Gemini and shows when your brand is mentioned.
- Review the list regularly and drop what is not working.
Check it worked by seeing your targeted keywords listed in a tracker with a starting position.
Tools for this job compared
| Tool | Good fit for | Pricing | Stated limits |
|---|---|---|---|
| Clearscope | Topic research and term guidance for SEO and content teams | Free trial, paid from $129/mo (checked 24 Oct 2026) | No free plan, only a 14 day free trial; caps on tracked queries, pages and drafts |
| SerpApi | Developers pulling structured SERP data | Freemium, paid from $25/mo (checked 24 Oct 2026) | Free plan has 250 searches a month and 50 searches per hour; pagination counts as separate searches |
| Serper | Google results, People Also Ask and related searches as JSON | Freemium | Tier prices and free query counts not stated publicly; one account at a time; data accuracy not warranted |
| Serplux | Topic ideas, competitor monitoring and daily position tracking | Pricing unknown | Prices not published; tracking described for Google only; vendor figures are its own claims |
| Lunroo | Keyword topics and briefs for content plans | Contact sales | Self-serve sign-up paused; accounts onboarded by email; pricing not published |
| Creaitor.ai | Keyword clustering plus AI-assistant visibility tracking | Paid | Starter tracks 1 domain; API access starts on Pro; dollar prices not shown for the first three plans |
| Search Atlas | Keyword tracking, audits and AI visibility in one account | Free trial, paid from $99/mo (checked 24 Oct 2026) | Starter has no LLM visibility projects or queries; extra sites are billed as add-ons |
Prices and plan limits change, so confirm the current terms on each vendor's site before you subscribe.
Common mistakes to avoid
- Trusting AI suggestions without checking the results. A suggested keyword may have no real demand or the wrong intent. Fix it by searching it and reading what ranks before you commit.
- Chasing volume alone. A big term may be out of reach or unrelated to what you sell. Fix it by scoring relevance and intent fit next to size.
- Mixing intents on one page. A single page rarely satisfies both a buyer and a researcher. Fix it by clustering by intent and giving each cluster its own page.
- Overlooking how the tool counts usage. API plans meter searches, and SerpApi counts pagination as separate searches. Fix it by estimating your request volume before you start.
- Doing it once. Rankings and demand shift. Fix it by tracking positions and reviewing the list on a schedule.
How we put this together
This is an evidence-based guide built from each tool's listing in our directory, which is compiled from the vendor's own site and pricing page. It is not a review based on using the tools. We don't rank tools, and nobody paid to be included. Some outbound links may be affiliate links; see our disclosure.
Frequently asked questions
Is there a free way to do keyword research with AI?
Partly. SerpApi has a Free plan with 250 searches a month, Clearscope has a 14 day free trial, and Search Atlas offers a 7-day trial. Serper says new users can start without a credit card. You can also read search results by hand for free.
Can AI do keyword research accurately?
AI is good at generating and grouping ideas, but it does not replace live data. Check intent against the real results and confirm any volume figures with a source you trust, since our listings do not describe volume for every tool.
How long does keyword research take?
It depends on how many topics you cover, and none of these listings gives a time. A full plan for a site takes longest, because you need to check the results for each promising keyword.
Does keyword research matter for AI search?
It still helps, though the way people ask is shifting. Clearscope's Query Tracking watches queries across ChatGPT and Gemini, and its query fan-out data shows the web searches AI platforms run to build answers. Use that alongside classic keyword work.
Which tool is good for developers?
SerpApi and Serper are APIs built for developers who want search results as structured data. SerpApi lists JSON and Markdown output, and Serper returns JSON. Check each vendor's terms before you build on them.
Compare more options in our Marketing and SEO category. Once you have keywords, see how to write SEO blog posts with AI.
