The model is only the beginning.
A useful model page should tell you where the weights come from, which license applies, how to run it, and what hardware was actually tested. YouRunAI brings those pieces together across text, coding, images, video, and agents.
Our evidence standard
This first catalog is based on linked publisher documentation. It is not a ranking of the newest or best models. No independent model benchmarks have been supplied or performed for this release.
When we add a test, the record includes its date, hardware, software, quantization, and context. Unknown fields stay marked “Not yet tested.” Publisher claims and our own measurements remain separate.
Local and cloud both belong here.
Local inference gives you direct control of your machine. Cloud compute opens up hardware you may not own. Neither is the universal answer. A documented self-hosting route does not guarantee a specific GPU or cloud deployment will work.
How the site may earn money
Some future compute links may be affiliate links, labeled where they appear. Provider recommendations must be explicitly configured. Pro is currently an interest list, and the public catalog is free to browse.
Help shape the field guide.
Join early access for future tests and practical setups, or tell us about a company use case.