Limits

AI Tokens has one tier. Every key gets the same treatment: there is no tier to choose when you create a key, no X-Siati-Tier header to send, and no multiplier on the prices.

• What you pay is the price of the model that answers, input and output apart, as listed in Pricing.
• How many requests a key may send: 120 per minute for all /v1 endpoints together, and the limit of each endpoint. The table, with what each 429 looks like, is in Rate limits.
• Which machine answers is our choice, among the machines that serve the model you ask for. The model id is the same on every machine.

The limits do not promise speed: a request that passes them may still wait in a machine's queue while it is busy. For capacity reserved to you, write to info@internetone.it.