Summary
get_provider_slugs is the small tool that saves the most time.
Every provider names the same model differently, and the failure is silent — a config that looks right and is rejected or, worse, quietly routes to a different model. Looking up the exact identifier one host expects is a two-second call that prevents a debugging session. The comparison tools are the strategic half: the same model's price can vary meaningfully by host, and this is where you find that before committing your inference budget.
What it is
A free, open endpoint over model pricing and benchmark data. It answered our probe with no credentials at all.
What you get
- get_all_models with filters for author, context size, throughput, time to first token and reasoning support
- get_model and search_models for one model or a keyword
- compare_models side by side, and compare_providers for the same model across different hosts
- get_benchmarks — models ranked by a named benchmark, with a minimum score filter
- get_providers with model counts and price ranges, and get_provider_slugs for the exact model identifier a given provider expects
Requirements
Nothing. Free with no key and no signup; the site is funded by sponsorship rather than by the endpoint.
Setup effort
Paste a URL — add the endpoint to your client