MiniMax Speech 2.8 HD
Model Information
Display Name: MiniMax Speech 2.8 HD
API Model ID: minimax/speech-28-hd
Category: Text To Audio
Description: MiniMax Speech 2.8 HD is a studio-grade text-to-speech model with voice cloning, emotion control, and multilingual support across 32 languages. Use any system voice or a custom voice cloned via MiniMax Voice Cloning. ## Key Features - **Voice cloning support** — use a `voice_id` from `minimax/voice-cloning` to speak in any cloned voice - **Emotion control** — happy, calm, sad, angry, fearful, disgusted, surprised, or auto - **32 languages** — English, Spanish, French, German, Italian, Portuguese, Japanese, Korean, Hindi, Arabic, Chinese, and more - **Fine-grained control** — pitch (-12 to +12 semitones), speed (0.5x–2x), volume (0–10) - **Multiple formats** — MP3, WAV, FLAC, PCM with configurable sample rate and bitrate - **Long text** — up to 10,000 characters per request - **Custom pauses** — insert pauses with <#0.5#> markers in text ## Parameters | Parameter | Type | Required | Default | Description | |-----------|------|----------|---------|-------------| | text (prompt) | string | Yes | — | Text to narrate (max 10,000 chars) | | voice_id | string | No | English_Wiselady | System voice or cloned voice_id | | emotion | string | No | auto | happy, calm, sad, angry, fearful, disgusted, surprised | | speed | number | No | 1.0 | Speech speed (0.5–2.0) | | pitch | integer | No | 0 | Semitone offset (-12 to +12) | | volume | number | No | 1.0 | Loudness (0–10) | | audio_format | string | No | mp3 | Output format: mp3, wav, flac, pcm | | language_boost | string | No | auto | Language hint for better pronunciation | | sample_rate | integer | No | 32000 | Sample rate in Hz (8000–44100) | ## Preset Voices English_Wiselady, English_Deep-VoicedGentleman, English_SweetLady, English_ExcitedGirl, English_WarmBoy, English_NarrationMan, English_NarrationLady, English_CheerfulGirl, English_StorytellerMan, English_PassionateHero, and more. ## Usage with Cloned Voice 1. Clone a voice with `minimax/voice-cloning` → get `voice_id` 2. Use that `voice_id` here with any text + emotion combination
How to Use This Model
To use MiniMax Speech 2.8 HD via the HInow.ai API, use the model ID: minimax/speech-28-hd
API Request Example (Chat/Text)
POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json
{
"model": "minimax/speech-28-hd",
"messages": [
{"role": "user", "content": "Your message here"}
]
}
Pricing
- per_1000_chars: $0.30
Available Parameters
- voice_id: System voice name or cloned voice_id from minimax/voice-cloning
- emotion: Emotion style for the speech (Options: auto, happy, calm, sad, angry, fearful, disgusted, surprised)
- speed: Speech speed multiplier (0.5-2.0)
- pitch: Semitone offset (-12 to +12)
- volume: Loudness (0-10)
- audio_format: Output audio format (Options: mp3, wav, flac, pcm)
- language_boost: Language hint for better pronunciation
- sample_rate: Audio sample rate in Hz (8000-44100)
Quick Reference
To use this model, set: "model": "minimax/speech-28-hd"
Featured: No
Documentation: https://hinow.ai/models/minimax/speech-28-hd
API Endpoint: https://api.hinow.ai/v1
