Frequently Asked Questions

What is TokenFeeds?
A free hub that aggregates and compares pricing, latency, and throughput for compute and LLM tokens across providers.
Where does the data come from?
Public pricing from providers and aggregators (e.g. OpenRouter), plus performance metrics from third-party measurement (Artificial Analysis). We distinguish the venue (who bills you) from the underlying provider (who runs inference).
How often do prices update?
Price roll-ups refresh every minute; history is retained for the trend charts.
What's the difference between a venue and a provider?
The venue is the billing surface — the entity you pay. The underlying provider is the inference server actually running the model. The same model via an aggregator vs. a direct provider are different routes.
What does "best price" mean?
The lowest published list price among routes available to a default user; private, BYOK, and disabled routes are excluded.
How is speed measured?
Time-to-first-token (TTFT) and throughput come from third-party measurement, not from TokenFeeds itself.
Is TokenFeeds free?
Yes — browsing prices and performance is free.
How do I report an issue or a missing model?
Email support@tokenfeeds.com.