KWFinder, what the keyword tool returns, where the free lookup stops and the one job it is clearly the right buy for

KWFinder is the keyword research tool inside Mangools, and it does two things from one box. Search by keyword and it returns search volume with historical data, a keyword difficulty score, and lists of related and competitor keywords. Search by domain and it returns the keywords a site already ranks for, which is how most people actually use it. It is sold as part of a five tool package rather than on its own, and knowing what the other four are is half of deciding whether it fits.

Verified against Google's India geo target alone on 22 October 2026, kw finder draws 1,000 searches a month in India, which is a real and steady stream of people trying to work out whether this is the tool to start with. Everything below was read off mangools.com on 22 October 2026, HTTP 200, so the claims are checkable at the source rather than remembered.

The five tools, because you are buying all of them

Mangools publishes the package on its own navigation, and the split is clean. Each tool answers one question.

That last row is the one to read twice before buying. Rank tracking updates weekly on the ordinary plans and daily only on the Agency plan, which the vendor states plainly. If your reason for buying a suite is to watch a position move day by day, that is a tier decision rather than a feature decision, and it is better to know it now than in week two.

The published numbers, and what each one is telling you

The KWFinder page prints four figures about its own scale, read 22 October 2026.

• 2.5 billion related keywords in the database, described as growing by about 20 million a month.
• Around 100 million competitor keywords checked each month, which is a usage figure rather than an index figure.
• Around 65,000 supported locations, including cities and districts as well as countries.

The location count is the number that matters most to an Indian buyer and it is easy to skim past. Local keyword research falls apart when the only geography available is the country, because demand for a service in Indore and in Mumbai are different questions with different competitors. A tool offering district level locations can answer the question you actually have.

The vendor also states that it provides exact search volumes and the most accurate keyword difficulty. Treat both as the vendor's claim about its own product, which is what they are, and test them the way you would test any supplier claim. The test takes ten minutes and is described below.

Where the free lookup stops

Published on the same page, read 22 October 2026, the free allowance is specific and small. Five lookups per 24 hours, returning 15 related keywords and 5 competitor keywords per lookup. There is a sign up for a free account and a separate list of free mini tools that need no subscription at all.

Five lookups a day is enough to answer a question and not enough to plan a quarter, which is a reasonable place for a vendor to draw the line and an unreasonable place to run a project from. Two consequences follow.

• Spend your five lookups on domains, not keywords. A domain lookup returns a list of what a competitor already ranks for, which is far more information per lookup than a single keyword search, and it is the lookup you cannot easily reproduce elsewhere for free.
• Write down the allowance with the date you read it. Free tiers move. This page carries the date for exactly that reason, and so should your note. If you are comparing several tools on their free limits, we keep a page on where each free backlink tier stops which follows the same discipline on the link side of the question.

The free mini tools, which are the interesting part of the listing

Mangools publishes a set of free tools outside the subscription, and read on 22 October 2026 the list includes a SERP simulator, a SERP location changer, a Reddit threads finder, a SERP volatility checker, a browser extension, an SEO content optimizer, a share of search tool, and two tools aimed at assistants rather than search results: an AI Search Grader and an AI Search Watcher.

Those last two are worth naming plainly, because they are a rival measuring the same surface we do and a reader deserves to know they exist. The honest comparison is on scope rather than quality: these are free single checks beside a keyword suite, and the question a buyer should ask about any of them, ours included, is which assistants are covered, how often the reading is repeated, and whether the sources behind each answer are shown. We have written up what our own measurements found on exactly that surface, including that review directories took 0.8 per cent of citations across 597 readings of ChatGPT and Gemini, measured 15 October 2026.

The SERP simulator deserves its own mention because it is the one most people get value from on day one, and it has a job that no amount of keyword data does. We cover it separately in what each feature of the Mangools SERP simulator shows.

The input rules, which catch people out on the first search

The keyword field states its own constraints and they are tighter than most tools. A keyword must be 2 to 80 characters, a maximum of 10 words, and may contain only letters plus the characters and, hash, and dollar. Read 22 October 2026.

In practice that means a long natural language question with punctuation in it will be refused, and so will a keyword list pasted with commas. If you are researching the way people actually phrase a question out loud, which is increasingly how buying questions get asked, you will hit this limit quickly and will need to shorten the phrase before the tool will accept it.

The ten minute test, instead of reading reviews

Every keyword tool models volume and difficulty from its own data, so the only comparison that means anything is one you run on keywords where you already know the answer.

• Pick five queries you already rank for and have Search Console data on. Compare the tool's volume with your own impressions for those queries over the same period. They will not match, because impressions are not searches, but the ratio should be consistent across the five. An inconsistent ratio means the model is weak on your subject.
• Pick three queries you tried and failed to rank for. Read the difficulty score on each. If the tool scored them easy, its difficulty model does not describe your market, and that is the score you were going to make decisions on.
• Run one domain lookup on your closest competitor and check twenty of the returned keywords by hand in a clean browser. The share that are genuinely ranking is the tool's accuracy on the one job you will use it for most.
• Run one lookup with a city level location if local demand matters to you, and check that the returned SERP matches what you see from that city. This is where a tool with 65,000 locations should separate itself from one with a country list.

Do that before the trial ends and the decision makes itself, which is a better use of an afternoon than reading comparison tables written by affiliates.

The job it is clearly the right buy for

A small team or a single practitioner who needs keyword research, a rank check and a basic link look from one subscription, and who does not need daily rank data. The suite is coherent, the tools each do one thing, and the interface is the thing users consistently praise, which matters more than feature count when the tool has to be used weekly by somebody whose job is not SEO.

It is the wrong buy for three cases. A team needing daily rank movement on ordinary tiers, because that is an Agency plan feature by the vendor's own description. A team whose main constraint is backlink index size, where the specialist link tools are built around a published index figure and should be compared on that. And an enterprise crawl of a very large site, which is a crawler problem rather than a keyword problem.

Questions about KWFinder and the Mangools suite

Is KWFinder available on its own

It is presented as the keyword research tool within the Mangools package alongside LinkMiner, SERPChecker, SiteProfiler and SERPWatcher. Plan what you will do with the other four before you judge the value, because you are buying the set.

What does the free version actually allow

Five lookups per 24 hours, with 15 related keywords and 5 competitor keywords returned per lookup, read on the vendor's page on 22 October 2026. Spend them on domain lookups rather than single keywords and you will learn considerably more per lookup.

How accurate are the search volumes

Every tool reports a model, and the vendor describes its volumes as exact. The useful move is to test it against five queries you already have Search Console impressions for, because that comparison is specific to your own subject and your own market in a way that no published accuracy claim can be.

Does it work for Indian local keyword research

It publishes around 65,000 supported locations covering cities and districts as well as countries, read 22 October 2026, so city level research is in scope. Verify on one keyword you know well, by comparing the tool's local SERP against a clean browser search from that city.

What is keyword difficulty actually built from

Difficulty scores in every tool of this kind are built from the link strength of the pages currently ranking, which is why two tools disagree about the same keyword: they disagree about the link graph first. Treat a difficulty figure as a sorting device across a list, not as a threshold that decides whether to publish.

Is a keyword suite still the right first purchase

For search results it is, because volume and difficulty are the two inputs that stop a content plan being a guess. Pair it with Search Console, which is free, reports your own real impressions and positions, and is the only source in the stack that is measuring rather than modelling.

The lookup this suite does not have a field for

Every number in KWFinder starts from the premise that somebody types a phrase and sees a list of links. The phrase is the unit, which is why the input field can cap you at ten words. A growing share of buying questions is not a phrase at all. It is a sentence spoken to an assistant, answered once, naming two or three companies, with no volume figure behind it and no difficulty score to look up, because nobody is publishing a keyword database for questions nobody types into a search box. Mangools has clearly noticed, which is why there is an AI Search Grader in its free list. The reading we take at AI Knows Us is the repeated one across ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, dated each time, so the answer to whether you are named stops being an impression and becomes a series.