> ## Documentation Index
> Fetch the complete documentation index at: https://docs.comfy.org/llms.txt
> Use this file to discover all available pages before exploring further.

# Breeze TTS 2 in ComfyUI: Text to Speech and Voice Tools

> Turn a script into speech in ComfyUI with Breeze TTS 2: native English and Chinese, 51 languages, voice cloning, voice design, remix and localization.

**Breeze TTS 2** is Breeze Blue's open-weight voice model for text to speech, with native English and Chinese and a multilingual variant that covers 51 languages.

In ComfyUI it runs through partner nodes. **Breeze Text to Speech** renders a script in a chosen voice. **Breeze Voice Selector** picks a featured voice, or any voice id you own. Four more nodes create and adapt voices: **Breeze Voice Clone** copies a voice from a recording, **Breeze Voice Design** invents one from a description, **Breeze Voice Remix** shifts an existing voice, and **Breeze Voice Localize** moves a voice into another language. All are cloud API nodes: they need a Comfy account with credits and download no local model. Rates are on the [partner node pricing page](https://docs.comfy.org/tutorials/partner-nodes/pricing.md).

> [!TIP]
>
> To use the Partner Nodes, you need to ensure that you are logged in properly and using a permitted network environment. Please refer to the [Partner Nodes Overview](https://docs.comfy.org/tutorials/partner-nodes/overview.md) section of the documentation to understand the specific requirements for using the Partner Nodes.

> [!TIP]
>
> > [tab] Local users
> >
> > Make sure your ComfyUI is updated.
> >
> > * [Download ComfyUI](https://www.comfy.org/download)
> > * [Update Guide](https://docs.comfy.org/installation/update_comfyui.md)
> >
> > Workflows in this guide can be found in the [Workflow Templates](https://docs.comfy.org/interface/features/template.md).
> > If you can't find them in the template, your ComfyUI may be outdated.
> >
> > If nodes are missing when loading a workflow, possible reasons:
> >
> > 1. You are not using the latest ComfyUI version (Nightly version)
> > 2. Some nodes failed to import at startup
>
> > [tab] Cloud users
> >
> > * [Cloud](https://cloud.comfy.org) will update after ComfyUI stable release.
> >
> > So, if you find any core node missing in this document, it might be because the new core nodes have not yet been released in the latest stable version. Please wait for the next stable release.

## Text to speech

Write the line, and the node returns audio with it spoken in the selected voice. Audio tags such as `(sighs)` or `(laughs)` are performed rather than read aloud, so a pause or a laugh comes out of the model instead of off the page.

The template speaks in a cloned voice. `breeze_tts2_tts_clone_audio.wav` runs through **Breeze Voice Clone**, and the returned voice id feeds **Breeze Voice Selector** with `language` set to `custom`, which drives **Breeze Text to Speech**. To use a featured voice instead, connect the other **Breeze Voice Selector** in the template (language `English`, voice `Elizabeth`) to the `voice` input of **Breeze Text to Speech**; it ships in the workflow but is not connected.

![Breeze TTS 2 text to speech workflow preview](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/main/templates/breeze_tts2_tts-1.webp)

> [card] [Run on Comfy Cloud](https://cloud.comfy.org/?template=breeze_tts2_tts\&utm_source=docs\&utm_medium=referral\&utm_campaign=breeze-tts-2)
>
> Open in Comfy Cloud

> [card] [Download Workflow](https://github.com/Comfy-Org/workflow_templates/blob/main/templates/breeze_tts2_tts.json)
>
> Download JSON or search "Breeze TTS 2: Text to Speech" in Template Library

Two more groups ship bypassed: **Breeze Voice Remix** changes a voice you already have, and **Breeze Voice Design** builds one from a description. Both return a voice id on their `voice` output, which ships unconnected. Connect that output to a **Preview as Text** node to read the id, or straight to the `voice` input of **Breeze Text to Speech**. Select the group's nodes, press **Ctrl+B** to clear the bypass, run the graph, then paste the id into **Breeze Voice Selector** with `language` set to `custom`.

**Input material**

> [card] [breeze\_tts2\_tts\_clone\_audio.wav](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/main/input/breeze_tts2_tts_clone_audio.wav)
>
> Load into the `Breeze Voice Clone` node, which the template already wires to the speech node

## Delivery control

`delivery` sets how the line is read:

* `neutral` keeps the voice close to its own sound. `speaker_boost` holds it closer to the original.
* `directed` follows written `instructions`, such as "warm and unhurried, smiling on the last word". Write them in the same language as the text, and set `strength` to control how literally the voice follows them.

`speed` changes the tempo from 0.5 to 2.0 without changing the pitch.

## Making voices

| Node | Input | What you get |
| - | - | - |
| Breeze Voice Clone | One speaker, at least 3 seconds. Only the first 60 seconds of the file are read, and Breeze keeps the best 30 seconds. `remove_background_noise` helps on a rough phone take | A new voice id, plus an audio preview |
| Breeze Voice Design | `description` of 3 to 500 characters (age, gender, accent, tone, pace) and a `preview_text` line. The voice is built from that line, so write it in the target language | A new voice id, plus an audio preview |
| Breeze Voice Remix | An existing voice, and either a `change` described in words or one of 27 presets across age, gender, pitch, pace, energy, emotion, role-play, and delivery. `strength` of 1 to 10 sets how far it moves | A changed copy of the voice |
| Breeze Voice Localize | An existing voice and a target language | The same voice speaking another language |

Remix keeps the original language; use Localize to cross languages. For cloning, only use a voice you have the right to reproduce: consent is required.

## Node controls

| Control | Values | Applies to | Effect |
| - | - | - | - |
| `voice` | featured voices, or a custom id | Text to Speech | The voice to speak with, from Breeze Voice Selector or another voice-producing node |
| `text` | up to 1000 characters | Text to Speech | The line to speak. Audio tags in parentheses are performed |
| `language` | eight featured languages, or `custom` | Voice Selector | Groups the featured voices. Languages without a featured voice need a voice id in `custom`, such as one made with Breeze Voice Localize |
| `language` | 51 languages | Text to Speech, Design | Language of the text. English and Chinese can use either model; every other language uses the multilingual one |
| `model` | `Breeze TTS 2`, `Breeze TTS 2 Multilingual` | Text to Speech, Design | Only shown for English and Chinese, which run on both |
| `delivery` | `neutral`, `directed` | Text to Speech | Plain reading, or reading that follows written instructions |
| `speed` | 0.5 to 2.0 | Text to Speech | Tempo, with the pitch unchanged |
| `seed` | integer | all voice nodes | Controls whether the node re-runs; results vary between runs |

A voice sounds best in its own language. To speak another language in a familiar voice, run it through Breeze Voice Localize first.

## Get started

1. Update ComfyUI to the latest version
2. Go to Template Library, search for `Breeze TTS 2`


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.