The Voho Speech API turns text into speech over HTTP or WebSocket. Every endpoint lives under one base URL and takes the same bearer token.

​
Authentication

Every request needs an API token in the Authorization header.
Tokens are created in the Voho console and begin with voho_sk_live_.
The plaintext token is shown once, at creation. Only its hash is stored, so a lost token cannot be recovered — create a new one and revoke the old. Revoking takes effect immediately.
A missing, malformed, or revoked token returns 401:

​
Endpoints

​
Response headers

Every response carries X-Voho-Engine. Successful synthesis calls add the details of what was produced:
X-Voho-Characters and X-Voho-Cost-Cents let you track spend per request without a separate usage call. Log them alongside your own request IDs.

​
Credits

Synthesis draws down a credit balance. Pricing is per 1,000 characters and depends on the model: Cost is rounded up, so a very short request still costs at least one cent. When the balance reaches zero, synthesis returns 402:
Streaming calls are billed after the stream finishes, so a synthesis that fails partway is not charged.

​
Errors

Errors are JSON with a stable code and a human-readable message. Branch on code, never on message.
See Errors for the full list.

​
Limits

  • 5,000 characters per request. Longer text returns text_too_long — split it into sentences and concatenate the audio.

Synthesise speech

The core endpoint.

Browse voices

Find a voice by dialect.