MiniMax Speech 2.8 HD API reference

Call MiniMax Speech 2.8 HD from your own code with one PopNow API key. Below are the exact model IDs to send, every input field the model accepts, what each request costs in credits, and a request you can paste into a terminal.

At a glance

Made by: MiniMax
Output: Audio
From: 18 credits per 1,000 characters ($0.18)
Typical time: about 9 s
Model IDs: minimax/speech-2.8-hd/text-to-speech

1 credit = $0.01. A task holds the highest possible charge when it is accepted and settles the real amount when it finishes, rounded up to whole credits; failed tasks are not charged. Ask POST /api/v1/jobs/quote for the exact figure before you run.

Text to Speech

minimax/speech-2.8-hd/text-to-speech

Price

Option	Credits (per 1,000 characters)	US dollars
`Default`	18	$0.18

Parameters

Field	Type	Default	Allowed	What it does
`text`Required	`string`	—	up to 5,000 characters	The words to be spoken, exactly as written. Sound tags such as (sighs) or (laughs) are performed. Billed per 1,000 characters.
`voice_id`Optional	`string`	`Calm_Woman`	Wise_Woman, Friendly_Person, Inspirational_girl, Deep_Voice_Man, Calm_Woman, Casual_Guy, Lively_Girl, Patient_Man, Young_Knight, Determined_Man, Lovely_Girl, Decent_Boy, Imposing_Manner, Elegant_Man, … (+458)	System voice. The label shows the voice name and, where known, the language it is built for.
`emotion`Optional	`string`	`neutral`	neutral, happy, sad, angry, fearful, disgusted, surprised	Emotion of the reading.
`speed`Optional	`number`	`1`	0.5 to 2	Speaking rate; 1 is the voice's normal pace.
`pitch`Optional	`integer`	`0`	-12 to 12	Pitch shift in semitones; 0 keeps the voice's own pitch.
`volume`Optional	`number`	`1`	0.1 to 10	Loudness multiplier; 1 is the normal level.
`language`Optional	`string`	`auto`	auto, zh, yue, en, ar, ru, es, fr, pt, de, tr, nl, uk, vi, … (+27)	Language hint that improves pronunciation; auto detects it from the text.
`output_format`Optional	`string`	`mp3`	mp3, wav, flac, pcm	File format of the result; pcm is headerless 16-bit little-endian samples.
`sample_rate`Optional	`integer`	—	8000, 16000, 22050, 24000, 32000, 44100	Sample rate in Hz. The model uses 32000 when omitted.
`bitrate`Optional	`integer`	—	32000, 64000, 128000, 256000	MP3 bitrate in bits per second; ignored for other formats. The model uses 128000 when omitted.
`audio_channels`Optional	`string`	—	mono, stereo	Mono or stereo output. Mono when omitted.
`english_normalization`Optional	`boolean`	`false`	—	Read numbers, dates and abbreviations in English text the way a person would say them.
`pronunciation_dict`Optional	`array<object>`	—	up to 100 items	Custom pronunciations: each alias in the text is read as its pronunciation.

Example request

This request runs as it is; any field you leave out uses its default. Put your key in POPNOW_API_KEY first.

POST /api/v1/jobs/createTask?wait=60

curl -X POST "https://api.popnow.ai/api/v1/jobs/createTask?wait=60" \
  -H "Authorization: Bearer $POPNOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
  "model": "minimax/speech-2.8-hd/text-to-speech",
  "input": {
    "text": "Hi there! This voice was created with the PopNow API in just a few seconds."
  }
}'

The response carries a taskId straight away. Read the result with GET /api/v1/jobs/recordInfo, or add callBackUrl to be notified.

GET /api/v1/jobs/recordInfo

curl "https://api.popnow.ai/api/v1/jobs/recordInfo?taskId=TASK_ID_FROM_CREATE_TASK" \
  -H "Authorization: Bearer $POPNOW_API_KEY"