# GLM Text to Speech - Documentation

> For the complete documentation index, see [llms.txt](/llms.txt). Markdown is available with `Accept: text/markdown` and `.md` URL variants.

Source: /docs/api-reference/model-apis-glm-tts

# GLM Text to Speech

POST

/

v3

/

glm-tts

GLM Text to Speech

cURL

```
curl --request POST \
--url https://api.novita.ai/v3/glm-tts \
--header 'Authorization: <authorization>' \
--header 'Content-Type: <content-type>' \
--data '
{
"input": "<string>",
"speed": 123,
"voice": "<string>",
"volume": 123,
"response_format": "<string>",
"watermark_enabled": true
}
'
```

Convert text to natural speech using GLM-TTS, supporting multiple voices, emotion control, and tone adjustment.

##

[​](#request-headers)

Request Headers

[​](#param-content-type)

Content-Type

string

required

Supports: `application/json`

[​](#param-authorization)

Authorization

string

required

Bearer authentication format, for example: Bearer {{API Key}}.

##

[​](#request-body)

Request Body

[​](#param-input)

input

string

required

The text to convert to speechLength limit: 0 - 1024

[​](#param-speed)

speed

number

default:1

Speech speed, default is 1.0, range [0.5, 2]Value range: [0.5, 2]

[​](#param-voice)

voice

string

default:"tongtong"

required

The voice to use for audio generation, supporting both system voices and cloned voices. System voices include: tongtong (Tongtong, default voice), chuichui (Chuichui), xiaochen (Xiaochen), jam (Dongdong Zoo jam voice), kazi (Dongdong Zoo kazi voice), douji (Dongdong Zoo douji voice), luodo (Dongdong Zoo luodo voice)

[​](#param-volume)

volume

number

default:1

Volume, default is 1.0, range (0, 10]Value range: [0, 10]

[​](#param-response-format)

response_format

string

default:"pcm"

Audio output format, defaults to pcm formatOptional values: `wav`, `pcm`

[​](#param-watermark-enabled)

watermark_enabled

boolean

Controls whether to add watermark when generating AI audio. true: Enables explicit watermark and implicit digital watermark for AI-generated content by default, complying with policy requirements. false: Disables all watermarks, only effective for users who have completed watermark removal action.

##

[​](#response)

Response

Request processed successfully, recommended sample rate is 24000
Format: `binary`

Last modified on December 19, 2025