Skip to main content
Rime is a speech synthesis platform built for conversational applications. In SentiVue Talk it provides the voice that speaks your agent’s responses.

​
What you need

​
Getting your API key

  1. Sign in to the Rime dashboard.
  2. Open API Tokens.
  3. Create a token and copy its value.
Copy the key and store it somewhere safe before leaving the page. A Rime key can synthesize speech against your account, so treat it like a password.

​
Adding the connection

  1. Go to Connections and choose Rime.
  2. Paste the API key.
  3. Save.
Saving stores the key. It is not checked against Rime at that moment, so a wrong or expired key first shows up as a silent agent on a live call — place a test call after saving.

​
What you can select afterwards

Choose your voice on Model Configurations. Model — enter the model id, such as coda. See Rime’s models page. Voice — a voice is identified by its speaker name, for example astra. Browse the Coda voices and Mist v3 voices catalogues. Language — three-letter ISO 639-2 codes such as eng or spa. A voice belongs to one language, so the code must match the voice you picked.

​
Completing the pipeline

Rime drives voice only. You also need: Each is a separate connection.

​
Troubleshooting

The key is only exercised when a call runs. Confirm you copied the full token from the API Tokens page in the Rime dashboard, then place another test call.
Rime reports a lapsed subscription the same way it reports a bad key, so a billing problem looks identical to a credential problem. Check the subscription status on your Rime account before regenerating the token.
Voices do not cross languages on Rime, and no voice is available on every model. Confirm the speaker name, the model and the language code are a combination listed together in the voice catalogue.
Rime rejects connections opened too quickly. It does not publish a concurrency figure, so if you run many simultaneous calls, ask Rime support what your account supports rather than inferring it from what you see.