StepAudio 2.5 TTS | EmpirioLabs AI Docs

StepFun · Audio Generation

POST /v1/audio/speech

Contextual StepFun text-to-speech model with natural-language voice direction and expressive delivery.

At a glance

Field	Value
Model id	`stepaudio-2-5-tts`
Model release date	2026-04-21
Input modalities	Text
Output modalities	Audio
Context window	-
Weight precision	-
Region	International
Features	text_to_speech, contextual_tts, voice_control, multilingual
Native inference	No
New	Yes
Supported endpoints	`POST /v1/audio/speech`

Pricing

Charge	Spec	Rate
Synthesis	per 10,000 characters	$0.85

Example request

$ curl https://api.empiriolabs.ai/v1/audio/speech \
>   -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \
>   -H 'Content-Type: application/json' \
>   -d '{"model": "stepaudio-2-5-tts", "input": "Hello from EmpirioLabs."}'

Parameters

Parameter	Type	Required	Default	Description
`input`	string	yes	-	Text to synthesize. Maximum 1,000 characters. · Max: 1000
`voice`	string	no	`"lively-girl"`	Official voice ID or custom cloned voice ID.
`response_format`	enum	no	`"mp3"`	Output audio format. · Allowed: `mp3`, `wav`, `flac`, `opus`, `pcm`
`speed`	number	no	`1.0`	Speech speed. · Range: 0.5 – 2
`volume`	number	no	`1.0`	Output volume. · Range: 0.1 – 2
`sample_rate`	enum	no	`24000`	Output sample rate in Hz. · Allowed: `8000`, `16000`, `22050`, `24000`, `48000`
`instruction`	string	no	-	Global emotion or style guidance. Only supported by StepAudio 2.5 TTS. · Max: 200
`pronunciation_map`	object	no	-	Optional pronunciation mapping.
`markdown_filter`	boolean	no	false	Filter markdown syntax before synthesis.

Notes

Maximum input is 1,000 characters. Content in parentheses is treated as direction and is not spoken. Use instruction for global emotion or style guidance. StepAudio 2.5 TTS does not support voice_label.

Machine-readable schema: GET https://api.empiriolabs.ai/v1/models/stepaudio-2-5-tts.