Sign inJoin the waitlist

Free while we’re in beta.

Invited workspaces use the studio and the API at no charge. There are no paid plans, and no payment method to add.

Beta

$0

Studio, saved voices, live and streaming speech, batch speech, API keys and usage. Invite-only.

Join the waitlist

Beta limits.

Usage in the beta can be limited per workspace or per API key, and any limit can change at any time without notice. The figures below are the ones the service enforces today, read from its configuration.

Speech, voices and keys

Characters per workspace per calendar month (UTC), studio and API together
500,000
Concurrent speech streams per API key, unless the key sets its own
2
Designed voices you haven’t saved are kept for
24 hours
Voices cloned from your own recordings
Not available
Limits on an API key’s concurrent streams and monthly characters
Set per key

Batch speech

Items per batch request
1,000
Characters per batch item
13,000
Batch request size
16 MiB
Batch results are kept for
24 hours

The beta is provided as it is.

We may change, pause or stop the service, or any part of it, at any time. A request over a limit, or beyond the capacity the service has at that moment, is refused with HTTP 429. Keep your own copies of anything you need, and don’t build on the beta where an outage would cause harm.

How characters are counted.

Usage is counted in input characters, when a request is admitted, and reported by workspace and API key.

  • Every character of the inputSpaces and punctuation included, markup excluded.
  • Voice design counts each finished voice’s sample textSaving a voice adds nothing.
  • Failures aren’t billableNeither are requests disconnected before the first audio arrives.
  • Batch counts the same wayEach item’s input, as if it were sent on its own.

Questions.

Is there a paid plan?

No. There are no paid plans, and no payment method to add.

How do I get an invite?

Join the waitlist below to ask for one. The beta is invite-only.

Can I clone a voice from my own recording?

Not on the service today. Voices come from descriptions: the studio’s directions, or voice design through the API.

Where can I see my usage?

The API’s usage endpoint reports your workspace’s characters, audio seconds, requests and errors, by day, key, model or language.

Join the waitlist.

The beta is invite-only. Leave your details to ask for an invitation.

The waitlist isn’t open.

Sign-ups haven’t started, so there is no form here to fill in, and this site collects nothing.