Compare 500+ AI models, side by side
Language models plus expert models for documents, speech, vision, translation and the web: every provider's price, capabilities and regions, and live speed for language models. Pick a model, then call it, or any other, with one API key.
Browse the catalog
Language models 500+ models
Chat, reasoning, coding and multimodal models from every major lab and open-weight family. Click a row to see every provider, price and live performance figure.
Expert models
Each row is one model for one task. Click it to open the model's page: every task it can do, with prices, parameters and a code sample.
Prices are compared within each task. Pick a task to compare its models directly.
Questions about the models
This page lists 750+ models: 500+ language models and 250+ expert models for documents, speech, vision, translation and the web. Every one of them is called with the same API key, so switching models means changing one parameter.
A model is one trained model, such as GPT-6.1 Sol. An endpoint is one provider's deployment of it, sometimes in a specific region. Popular models have several endpoints with different prices, context limits and hosting locations. Each model page lists all of them, cheapest first.
Yes. Use the EU hosting filter to see models with at least one endpoint in Europe, then pin that endpoint with its ID. Learn more about the EU endpoint.
They are measured on real Eden AI traffic over the last 30 days: median time to first token, latency, throughput and availability for each endpoint. Endpoints with too little traffic are not reported.
Yes. 10+ models currently have at least one free endpoint. Turn on the Free endpoint filter to list them.
One API key for every model
Route by price, speed or region, with automatic fallbacks. No monthly commitment, pay as you go.



.avif)






.png)


