Breeze Text to Speech - ComfyUI Built-in Node Documentation

Converts text into speech with Breeze TTS 2. Pick a Breeze voice, choose the language of the text, and optionally direct how the line is performed. English and Chinese can use either Breeze TTS 2 model; every other language uses Breeze TTS 2 Multilingual.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`voice`	Breeze voice to speak with, from Breeze Voice Selector or one of the Breeze voice creation nodes.	BREEZE_VOICE	Yes	-
`text`	Text to speak, 1 to 1000 characters. Audio tags such as `(laughs)` or `(sighs)` are performed instead of read. Default: empty.	STRING	Yes	1 to 1000 characters
`language`	Language of the text. English and Chinese can use either model; other languages use Breeze TTS 2 Multilingual. A voice sounds best in its own language, so for another language use a copy made with Breeze Voice Localize.	DYNAMIC_COMBO	Yes	Multiple options available
`delivery`	How the line is performed: `neutral` keeps the voice as it is, `directed` follows written instructions. Default: `"neutral"`.	DYNAMIC_COMBO	Yes	`"neutral"` `"directed"`
`speed`	Tempo multiplier that leaves the pitch unchanged. Default: 1.0.	FLOAT	No	0.5 to 2.0 (step 0.05)
`seed`	Controls whether the node re-runs; results are non-deterministic regardless of the seed. Default: 42.	INT	No	0 to 2147483647

Language “English” and “Chinese” Inputs

These appear only when language is English or Chinese.

Parameter	Description	Data Type	Required	Range
`model`	`Breeze TTS 2` or `Breeze TTS 2 Multilingual`. Default: `"Breeze TTS 2"`.	COMBO	Yes	`"Breeze TTS 2"` `"Breeze TTS 2 Multilingual"`

Delivery “neutral” Inputs

Parameter	Description	Data Type	Required	Range
`speaker_boost`	Keeps the voice closer to its original sound. Default: false.	BOOLEAN	No	true false

Delivery “directed” Inputs

Parameter	Description	Data Type	Required	Range
`instructions`	How to perform the text, written in the same language as the text. Default: empty.	STRING	No	Multiline text
`strength`	How strongly the voice follows the instructions. Default: 4.0.	FLOAT	No	1.0 to 10.0 (step 0.1)

Outputs

Output Name	Description	Data Type
`audio`	The generated speech as WAV audio.	AUDIO

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): e5e5201e8c951a878ef9e764db31e494285ddded6ec66d8e818ec132342bbe5c