Gemini Text to Speech
Gemini Text to Speech reads out the text you give it with one of the thirty voices of Google’s Gemini and returns a WAV audio file. Every voice speaks every language: the text sets the language. You can describe the tone and pace in words. If the text is long, it splits it at the end of sentences, reads each part and joins them into a single file.
You need a Google Gemini credential with your Google AI key, the same one the Gemini model uses. See Credentials. The text is sent to Google to produce the audio.
This is not the voice node for calls: to speak to a caller during a live call, the Voice Agent uses the Gemini TTS node. This component returns a finished file.
When to use it
- Produce voice-overs in a particular tone, for example warm and unhurried.
- Read texts in any language with the same voice.
- Let an agent answer with audio, using the component as a tool.
- Produce the audio for a list of texts in one go with Batch Mode.
It only produces WAV. If you need an MP3, use OpenAI Text to Speech or Inworld Text to Speech.
Configuration
| Field | What it does |
|---|---|
| API Key (required) | Your Google AI key. The key is filled in for you when you choose your Google Gemini credential. |
| Text to Speech (required) | The text to read out. You can type it or connect it from another component. |
| Voice | The voice. Default: Kore. There are thirty, such as Puck, Zephyr, Charon, Leda or Aoede. Every voice speaks every language. |
| Model | The voice model. Default: gemini-3.8-flash-tts, the most expressive. gemini-3.8-flash-lite-tts is cheaper. gemini-3.1-flash-tts-preview and gemini-2.5-pro-preview-tts are also available. |
Advanced options
| Field | What it does |
|---|---|
| Voice style | Optional. How you want it to sound, in words: “warm and unhurried”, “brisk and professional”. This is also where you set the pace: there is no speed field. |
Output
The Data output returns a WAV audio file (24 kHz, mono) with the text read out. Its name is the first words of the text, without accents or punctuation (for example, Hola_buenos_dias_Su.wav).
Connect it to a component that stores or sends files. An agent can also use the component as a tool to create audio when it needs to; see what tools are in Key concepts.
Example
A hotel prepares a welcome audio for each guest in their own language. An agent translates the welcome text into the booking’s language, Gemini Text to Speech reads it with the Aoede voice and Voice style set to “warm and calm”, and the flow sends the audio on the day of arrival.
Related components
- Speech to text and text to speech: every component that transcribes audio or produces it.
- Gemini Speech to Text: the opposite, turns audio into text with the same credential.
- OpenAI Text to Speech: another way to describe the tone in words, with MP3 output.
Tips
- In Voice style, describe only how it sounds, not the message: the text to read always goes in Text to Speech.
- It is the slowest of the five new text-to-speech components. Measured in our test environment on 3 November 2026: the WAV of the sentence “Hola, buenos días. Su cita está confirmada para el martes a las diez y media.” was produced in 3.3 s.