EvalAITools Check a repo

About EvalAITools

I’m El Mehdi, and I run EvalAITools on my own. Too many “best AI tools” pages repeat vendor launch posts, rarely say what something actually costs once you use it, and almost never admit where a tool falls short. This site tries to do the opposite.

So this site has one rule: if I can check something myself, I check it, and I show you how. If I can’t, I say so and point you to the source.

What you’ll find here

  • Measurements I can actually run. I clone public code repositories and count what’s really happening in them, like which agent frameworks are still maintained or how often commits credit an AI coding tool. I also cost out real workloads using official price lists, because a per-token price on its own tells you very little.
  • Comparisons with a verdict. I don’t think “it depends” is a useful answer on its own. When I compare tools, I tell you which one I’d pick for which job, and why.
  • Explainers on new models and products, built on primary sources and clearly separated from my own opinion.
  • A live dashboard. The homepage tracks more than 100 open-source AI tools from GitHub with a health score for each, the prices, status and outages of paid AI subscriptions, and API prices for hundreds of models. It updates on its own; how each number is collected is explained in full.

Articles that include my own measurements carry a yellow Includes our measurements marker, and each measurement section starts with a short note explaining exactly what was done, when and with what limits.

What I don’t do

  • I don’t claim to have tested a product I haven’t used. Most AI products I write about need paid business accounts I don’t have, and when that’s the case, the article says so.
  • I don’t invent results, user stories or “I spent 30 days with” anecdotes.
  • I don’t take payment for rankings or positive coverage. If that ever changes for a specific article, the article will say so at the top.

How I use AI to run this site

I’ll be direct about this, because it’s a site about AI tools and you should know how it’s made. I use AI assistants, including Claude, to help me research, write and run the scripts behind my measurements. I decide what gets covered, what the verdict is and what gets published, and I’m responsible for every claim on the site. When numbers come from a script, the article explains the method so you can repeat it.

Where to start

Corrections and contact

AI products change every week, so I will get things wrong or out of date. If you spot a mistake, email me at contact@evalaitools.com with the link and what’s wrong. I fix confirmed errors and update the date on the article. You can also read my editorial policy and how I review AI tools.