Pricing

Prices in EUR. The tier of the request multiplies the prices counted in tokens; audio, pictures and video have a fixed price per unit.

Chat models

Chat completions, completions, the Responses API and the answers of knowledge bases are counted in tokens, per model, with input and output counted separately: generating tokens costs far more than reading them. EUR per million tokens.

Input tokens cost the model's list price on every hardware. Output tokens cost the list price times the factor of the hardware that answers. Write model@hardware to choose the machine and pay its output price; with the model's name alone the service picks a machine that serves the model, at that machine's price. Each row below is one model on one kind of hardware.

A long prompt pays its input more, on the whole input of that request: over 32,768 tokens ×1.5, over 131,072 tokens ×2. The input prices below are for prompts up to 32,768 tokens.

An image sent to a model that reads images becomes input tokens, at the input price of the model.

Other services

What is not charged

• A request refused before it runs: any 4xx, for a parameter, the key, a rate limit, credits, or a prompt refused by the content check.
• A request that fails on our side: a 502 or a 503, a stream that breaks before its first piece of text, a video job that fails, a question to a knowledge base whose documents could not be searched or that has no document ready.
• A picture or a video clip refused by the check after it was made.
• A silent WAV file, answered with empty text without asking the model.

What is charged although it is incomplete: an answer cut at the token cap, with the tokens it used; a stream you close early, or that breaks after its text has started, with the tokens the engine reported or, when it did not, an estimate of about four characters per token; and with n above 1, every answer.

How a request is counted

For a request with P prompt tokens and C completion tokens, on a model with input price in and output price out per million tokens:

cost = (P × in + C × out) / 1_000_000

Embeddings, rerank and indexing have input tokens only. Audio, pictures and video are counted in their own units, in the table above. What each request cost is in your usage records, in the dashboard.

Limits

One level for every key: there are no tiers to choose and no multiplier, and you pay the prices above. Every key has two limits, counted per minute: 120 requests on all /v1 endpoints together, and the limit of the endpoint it calls. The first one reached answers 429. More in Rate limits.

Dedicated capacity

For continuous workloads or guaranteed throughput we reserve capacity for your exclusive use, with absolute priority and no contention with other customers, for a monthly fee set on the model and capacity required.

Write to info@internetone.it for a quote.

Prices exclude VAT: Swiss VAT at 8.1% is added on the invoice from the day Internet One Srl is registered for VAT.