How We Find Tools

This page explains how Free AI For That ("FAFT", "we") finds the tools it considers listing. Finding a tool never lists it. Every tool goes through the checks in How we check free offers before it can appear, and a member of our team approves each new listing.

1. What we look for

AI tools that people can use free for at least a month, and that do a real job well. In particular:

• tools for tasks and categories where our listings are thin;
• new tools, and new or bigger free offers from existing ones (a free plan launched or enlarged, free during a beta, a paid plan given free for a set time, a project made open source);
• tools we don't list today whose free offer now qualifies;
• free alternatives to popular paid tools;
• new uses of AI that we don't cover yet.

2. How we decide where to look

We start each day from evidence, not habit:

• What visitors need: searches on our site that find nothing, the pages people visit and leave, the tools they click through to, save and rate, and questions our help chat couldn't answer. We use counts and totals only, never anyone's personal details.
• Where we're thin: tasks and categories with few tools.
• What's changing: kinds of tools and uses of AI that are growing.
• What people send us: suggestions through our Submit page, and makers' applications.

From this we set a few working ideas for the day, for example "searches for free logo makers find nothing", and look for tools that answer them. Later we check whether the tools we added were visited and used. Ideas that work get more attention; ideas that don't are dropped.

3. Where we look

Anywhere on the open web: search results, makers' own websites and announcements, open-source code repositories, public launch forums and newsletters. These are only leads. Every fact we publish comes from the maker's own pages (see Verification policy, section 3).

We don't:

• copy descriptions, screenshots or lists from other directories, or treat any directory as evidence;
• use a source whose terms don't allow it;
• sign up, log in, or get around paywalls, logins or bot protection;
• give a lead priority because a maker paid: an application is checked like any other lead, and payment never raises its place in the queue.

4. How we choose which tools to check

We weigh each candidate on:

1. Usefulness: does it do a real, common job well, especially one our visitors need?
2. Its free offer, as the maker states it: how much is free, for how long, and with what catches.
3. The maker's credibility: a named company or team, published terms and privacy policy, documentation, and signs of real users. Credibility matters, but we also want to help new and early makers whose tools already do enough to be useful. For a young company we read its privacy policy and terms ourselves, and make sure they don't break the law or our rules, for example by claiming rights over your content, selling your data, or allowing what our Sponsor policy excludes.
4. Momentum and novelty: how fast it is growing, and how new it or its free offer is.

We set a candidate aside when the maker's own pages show it doesn't qualify: not an AI tool, no free offer lasting at least a month, a demo you can't keep anything from, free only for a narrow group, or an excluded category (adult content, gambling, crypto schemes, impersonation). Waitlists, thin wrappers around another AI model, and copies of tools we already list aren't chosen.

We keep the mix balanced. Each day about half of the candidates fill gaps and half are new, with limits for each task, category and company, so no single area or maker dominates. Up to about 30 candidates a day go on to our checks.

5. What we keep

We keep a record of every candidate we look at: where we found it, what we saw, and why it was chosen or set aside. Nothing is deleted. A tool set aside is looked at again later, in case its offer changes. Each day's list is kept, so we can see which ideas and sources led to tools people use, and adjust.

6. Who decides

Finding and choosing candidates is done with the help of AI models, under these rules. Listing a tool is always decided by our checks and by a member of our team.