Text to speech pricing

Definition

Text-to-speech pricing is based on the amount of text you synthesize, billed per character, starting from $6 per 1M characters on Scale. It is a linear model: the cost tracks the volume of text, with no per-request surcharge and, within a model, no premium for a particular voice or language. Simba 3.0 and 3.2 bill $10/$8/$6 per 1M characters on Starter/Pro/Scale. Simba 3.5 lists at $20/$16/$12, billed at $10/$8/$6 until 21 Nov 2026.

How character pricing works

You are charged for the characters of input text you send to be spoken. Because the unit is characters, cost is easy to predict: count the text, apply the rate. A notification is fractions of a cent; a full audiobook of a few hundred thousand characters is a few dollars of synthesis.

Estimating your cost

Take the total characters you expect to synthesize in a period and multiply by the per-character rate. Within a model, choosing a different catalog voice or language does not change the price, so your estimate depends on volume and the model, not on which voice you pick. Simba 3.0 and 3.2 bill $10/$8/$6 per 1M characters on Starter/Pro/Scale. Simba 3.5 lists at $20/$16/$12, billed at $10/$8/$6 until 21 Nov 2026.

Files and streaming cost the same

Delivery mode does not affect price. Streaming and file synthesis of the same text cost the same, because billing is on input characters, not on how the audio is returned. See the pricing page for current plan rates.