FAQ
Short answers to the questions people ask first.
What is the difference between Preview and Answer?
Preview shows the routing decision only: every model's predicted odds of getting your question right, its estimated price, and which one would be called. No model is called, so it is free. Answer sends the question on to that model and streams the reply, with its real cost shown.
Why do I need an account for answers?
Answering calls a real model, which costs real money, so it has to be tied to an account that can hold a free allowance and a credit balance. A preview never needs an account.
Which model answers me?
The one the router predicts is the best trade between the odds of a right answer and the price, for that question. You can see the pick and the reasoning in a preview. You cannot override it, in the chat or over the API. See How it works.
What does it cost?
A preview is free. Answers have a free daily allowance and then draw on prepaid credit. The numbers are on the Pricing page.
Can I use it from my own tools?
Yes. Create a key in Settings and point any tool that speaks the OpenAI or Anthropic format at Fugal. The Quick start takes about two minutes.
Where are my chats stored?
Signed out, in your browser only: nothing leaves it. Signed in, they sync to your account so you can see them on other devices, and only you can read them. You can export or delete everything in Settings.
Are my chats used for training?
No, unless you turn that on yourself. The switch is off by default and lives in Settings. The Privacy Policy has the detail.
Who builds the router?
Miners on Bittensor compete to build the best one, and the best is paid for as long as nothing beats it. See Why a subnet and the Leaderboard.