Keyword Research
Find & prioritize SEO keywords
Try asking
About
Keyword Research turns a seed topic into a full content strategy — finding keyword opportunities, breaking them into page clusters, and scoring them so you know exactly what to publish first. Every result comes with intent classification, difficulty estimates, and a content brief so agents can act on the data immediately.
Most SEO workflows require switching between multiple tools just to go from idea to plan. This one discovers keywords, reads the SERP, groups a site architecture, and ranks the highest-leverage pages. Checked 2026-08-26 against this manifest.
What you can do
- find_keywords — discover keyword ideas for any topic with live monthly search volume, intent, and difficulty
- analyze_keyword — get a full breakdown of a single keyword including SERP shape, content brief, and title angles
- cluster_keywords — group a list of keywords into logical page clusters with recommended slugs
- score_opportunities — rank a keyword list by opportunity to separate quick wins from long-term bets
- check_rankings — observe current organic positions with locale, device scope, and date evidence
- compare_visibility — compare named domains inside the same current result sets
- serp_features — record present search features and attributable target ownership
Who it's for
Content marketers planning an editorial calendar. SEO managers building a keyword strategy for a new product or market. Founders who want to understand the search landscape before commissioning content. Product teams researching what terms competitors rank for.
How to use it
- Start with find_keywords and provide your seed topic — optionally add a business goal to bias results toward revenue or signups
- Pass the results to cluster_keywords to organize them into a content architecture with page slugs
- Use score_opportunities to rank the clusters by priority so you know what to build first
- Use analyze_keyword on your top targets to get detailed SERP analysis and a content brief
- Use check_rankings, compare_visibility, or serp_features for dated current measurement; historical change requires persisted snapshots
Getting started
Try find_keywords with "AI CRM software", a demo-request goal, and limit 12. Representative manifest-schema shape: topic, keyword_ideas, next_steps [redacted]. Modeled signals are not guaranteed; timing varies.
Research related keywords for a topic so you can spot live monthly…
Break down a keyword so you can see intent, SERP features, monthly…
Group keyword lists into clear page clusters so teams can turn raw…
Prioritize a keyword list so you can decide what to publish now, what…
Research related keywords for a topic so you can spot live monthly search volume, intent, and the best gaps to target.
Break down a keyword so you can see intent, SERP features, monthly search volume, difficulty, and the best page to create.
Group keyword lists into clear page clusters so teams can turn raw research into a sane content architecture.
Prioritize a keyword list so you can decide what to publish now, what to save for later, and what to skip.
Observe current organic positions for a domain and keyword set with locale, date, result-depth, and device-scope evidence.
Compare the target and named competitors only inside the same current observed keyword result sets and result depth.
Record current search-result features per keyword and attribute ownership only when the observed evidence supports it.
- Live search volume now comes from the requested country for every country Google serves, instead of US figures outside a short list. An unknown country code is now reported before any work starts.
- Keyword ideas and analysis now include live monthly search volume when a search-volume account is connected.
- Added current ranking, competitor visibility, and search-feature observations with explicit coverage and no invented historical change.
- Longer keyword reports and larger scoring runs now have room to finish instead of being cut off part-way through.
- Opportunity scoring no longer reports a confusing formatting error when a scoring run is interrupted — it now says the run needs retrying.
- Keyword analysis now uses Gemini 3.7 Flash by default.
- Fixed search-powered skills (find_keywords, analyze_keyword, score_opportunities) that were failing with a connection error
- Added subtitle, expanded description, and agent instructions
- Initial release
Powered by
Related
Guides
- How to Do Keyword Research in ChatGPT (Without Fake Volumes)ChatGPT can cluster ideas. It cannot see Google volume. Ahrefs says unprompted KD and volume are fabricated. Connect a keyword tool, then cluster.
- Ahrefs alternatives for occasional keyword researchCompare metered keyword research with Ahrefs Lite, Starter, and free options. Understand live-volume access and what each approach covers.
- Best MCP Tools for SEO (Honest List)ToolRouter SEO/GEO tools in chat vs Ahrefs and Semrush MCP. We win on one connector and auto-provision. They win on proprietary SEO data.
Questions
Can I research keywords from a single topic or business goal?
Yes. `find_keywords` starts from a seed topic and can be biased with a business goal such as demos, leads, or signups. You can also pass a site URL to check existing coverage.
How do I turn keyword ideas into a content plan?
Use `cluster_keywords` to group raw ideas into page clusters, priorities, and URL slugs. It is useful when you want an information architecture instead of a flat keyword list.
Can I score which keywords are worth targeting first?
Yes. `score_opportunities` ranks keywords by modeled demand, difficulty, and opportunity so you can separate quick wins from longer-term bets.
What does a deep keyword analysis include?
`analyze_keyword` gives you intent, SERP shape, modeled demand, difficulty, and a content brief with title angles for the keyword you are evaluating.