The intelligence layer for AI.

Why we built this

Most AI directories are scrapers with an affiliate link attached. They list thousands of tools, rank them by who paid, and tell you nothing about which one fits your job. We wanted the opposite: tools scored on the same criteria, with a plain answer to "which one should I use?"

The catalogue is not the product. Anyone can list software. What is hard, and what we spend our time on, is the layer on top: a score you can trust, a comparison that reaches a verdict, a model rating that says which one suits your job, and a guide that puts them in the right order. The directory is the database. The judgement is the point.

Who runs Solthar

Solthar is run by Manish. He publishes every blog post and news story on the site, and the corrections and questions sent from the contact page go straight to him.

How we score

Every tool gets a Solthar Score from 0 to 100. That number is our editorial judgement, not a benchmark and not a user poll. We publish it whatever it says, and we name what each tool is bad at as plainly as what it is good at.

For AI models we go further with Model DNA, a rating across eight axes that decide which model suits which job. We keep that separate from published benchmark scores, because one is our opinion and the other is measurement.

How we make money

Right now, we don't. Solthar takes no payment for placement, ranking or inclusion. If that changes we will label sponsored entries clearly and keep them out of the rankings. No tool has ever paid to appear on this site.

What we get wrong

Plenty. Scores are judgements and reasonable people disagree with ours. Pricing changes faster than we can track it. Some listings carry review counts and traffic figures we have not yet sourced, and where that is true we show nothing rather than invent a number.

If we have something wrong, tell us and we will fix it.