A directory of AI tool directories, ranked by whether they send traffic

Every list of AI tool directories has a column of traffic numbers in it. Almost none of those numbers come from the directories, and the ones that do are printed on the page that is selling you a listing.

So this list works differently. We opened every site below in August 2026 and recorded what its own pages say — price, review model, whether a backlink is required. There are no traffic estimates, because we cannot verify them. At the end there is the method for ranking these yourself in a month, which produces an answer that is correct for your product rather than approximately correct for the average one.

Ranked launch boards

A daily or weekly leaderboard, votes, a comment thread. The launch is the event; the listing afterwards is a byproduct.

Two Product Hunt rules that catch people out: company accounts are prohibited, and you "cannot ask people directly to upvote your product. Instead, ask them to visit and comment."

We run one of these too — weekly rather than daily, so a listing gets seven days rather than twenty-four hours. Free launches require our badge, $29 removes that, and the weekly boards are public.

Category directories

Permanent, organised by what a tool does. No launch day. The listing either ranks for a term for years or does nothing.

Two things worth knowing about this group. First, the AI-specific catalogues have largely moved to paid submission, and the prices are not small — $347 is a real marketing decision, not a formality. Second, AlternativeTo is crowdsourced rather than submission-driven, and it explicitly declines "AI wrappers" and apps indistinguishable from existing alternatives. It is also the one place that captures "[competitor] alternatives" searches, which is high-intent traffic.

Our catalogue sits in this group: free, reviewed by a person, browsable by category, and readable by agents through an API and MCP server with no key required for reads.

Startup directories

BetaList is the notable one, and the notable fact is that it has gone fully paid. Their FAQ now reads: "All submissions are paid. There is no free submission option." Prices appear only after you log in. Every listicle that still calls BetaList a free option is out of date.

Launching Next takes free submissions reviewed daily, with a $99 upgrade for consideration within one business day. Indie Hackers has a products database, though it functions more as a community than a directory.

Community surfaces

Not directories. The same approach will fail.

Hacker News — Show HN publishes its rules and enforces them. On topic: "things people can run on their computers or hold in their hands". Off topic: "blog posts, sign-up pages, newsletters, lists". The project must be "something you've worked on personally and which you're around to discuss", and you should "make it easy for users to try your thing out, ideally without barriers such as signups or emails". A gated signup page posted as a Show HN gets flagged, not ranked.

Subreddits and Discord servers in your product's actual domain have the highest variance of anything here. A good post in a niche community beats every directory on this page combined; a promotional one is removed in four minutes.

Review sites

G2 and Capterra are built around buyer reviews rather than discovery. Both are free to list — G2 says "Setting up your profile is free" and "G2 is not pay to play", with paid marketing products sold separately — and both review submissions manually before publishing.

They matter for B2B software with a sales cycle attached and are close to irrelevant for a $9-a-month consumer tool. The work is also different in kind: you are collecting reviews, not writing a listing.

Two things we found that are worth naming

The badge is the free tier's real price. Fazier and Startup Fame both give you a free listing only if you put their link on your homepage, and both sell the same listing without that condition for $29 and $19/month respectively. That is not a criticism — we do it too — but it is the trade, and it is worth seeing it stated rather than discovering it after submitting.

Two of the directories on every listicle are dead. aitoolhunt.com returns a 404 on every path we tried. insanelycooltools.com redirects to a GoDaddy parking page; the domain is for sale. Both still appear in current "best AI directories" posts, which tells you how many of those posts are compiled from other posts.

The "submit to 100 directories" services

These exist, they publish prices, and the market has converged remarkably tightly:

The last two have identical tier names, counts, prices and strike-through anchors, which is worth a moment's thought before buying from either.

Our view: the listings themselves are not the problem — most of these are free directories you could submit to yourself. The problem is that you get a hundred listings with identical scripted copy appearing in the same week, no UTM tags, and no way to know which one sent anyone. You have bought the activity, not the outcome.

Ranking them yourself, in a month

1. Tag every submission. A distinct URL per directory: ?utm_source=fazier&utm_medium=directory. Do it at submission time; you cannot retrofit it.

2. Record what you did. Directory, date, status, listing URL. Four columns, and it solves the "did I already submit here" problem six months from now.

3. Wait thirty days. Launch-day traffic tells you about launch day. Directory traffic is a slow trickle, and thirty days is the shortest window that distinguishes nobody from two visitors a day.

4. Read it by source, and look at signups, not sessions. A directory sending 400 visits that convert at nothing is worth less than one sending 30 that convert at ten percent.

5. Keep the shortlist. Three or four will account for nearly all of it. That list is the real output — measured, and specific to your product, which is more than any post like this one can give you.

For most products, most directories send almost nothing and one or two send a genuinely useful trickle. That is not a reason to skip them: a submission costs twenty minutes and a listing is permanent. It is a reason to stop treating a submission run as a growth strategy and start treating it as a cheap experiment you actually measure.

The preparation checklist turns those twenty minutes into five, and what the paid tiers actually buy covers the money side in more detail.