Giggy

Giggy is a text-to-speech service with a large public voice catalog, covering more than 80 languages, plus any voices you create in your own Giggy account. Giggy tuning is passed in synthesizer.options.

{
  "verb": "say",
  "text": "Thanks for calling, how can I help you today?",
  "synthesizer": {
    "vendor": "giggy",
    "voice": "38607161-601a-517f-96c3-8708d1975b9c",
    "options": {
      "speed": 1.1
    }
  }
}

You can also set these in the portal under Extra Options on the speech credential, in which case they apply to every request that uses that credential. Settings supplied on the verb take precedence.

Voices and languages

Voice ids are UUIDs. Each Giggy voice has its own language, and the voice decides the language it speaks in. jambonz does not send synthesizer.language to Giggy, because Giggy rejects a language that differs from the voice’s own. Pick a voice for the language you want.

The public catalog has about 24,500 voices, too many to list in full. The portal’s voice list shows the top-rated public voices for each language and gender. It also includes the voices you have created in your Giggy account, marked “(custom)” and listed under every language, since Giggy does not record a language for them. Any other public voice works too: put its id in voice.

Giggy serves a single model, giggyspeech, which is the default.

Options

Both options are optional. An option you don’t set is not sent, so Giggy’s own default applies.

Endpoint override

Giggy credentials have no API base URL field in the portal. If you need to point at a non-default endpoint, such as a proxy, set api_uri in Extra Options alongside the tuning fields:

{
  "api_uri": "my-proxy.example.com",
  "speed": 1.0
}

Leave it out to use Giggy’s standard endpoint (giggy.ai).

Streaming

Giggy supports TTS streaming, so it can be driven token by token from an LLM using tts:tokens, for example from the agent verb. Giggy has no connection that accepts text as it arrives, so jambonz collects the tokens into whole sentences and sends each sentence to Giggy as a separate request. The first audio therefore starts once the first sentence is complete. Splitting at sentence boundaries keeps the intonation natural.

On barge-in, jambonz cancels the request in progress. Giggy does not return word-level timestamps, so jambonz cannot trim the assistant’s turn to the words the caller actually heard.