tts-skill

MiniMax TTS Skill

This Skill provides a complete wrapper for the MiniMax TTS API, supporting text-to-speech, voice cloning, and voice design features.

Quick Start

1. Environment Configuration

Make sure the environment variables are set:

export MINIMAX_API_KEY="your-api-key"

For detailed configuration instructions, see setup.md

2. Using the Python Module

import sys
import os

# 获取 skill 目录路径
skill_dir = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, os.path.join(skill_dir, "assets"))

from minimax_tts import text_to_audio, list_voices, voice_clone, voice_design, play_audio

Feature Overview

Feature	Function	Description
Text-to-Speech	`text_to_audio()`	Convert text to a speech file
List Voices	`list_voices()`	Get the list of available voices
Voice Cloning	`voice_clone()`	Clone a voice based on an audio file
Voice Design	`voice_design()`	Generate a voice based on a text description
Play Audio	`play_audio()`	Play an audio file

Detailed Documentation

Environment Configuration - API Key and dependency installation
Text-to-Speech - Detailed explanation of TTS features
Voice List - Available voices and filtering
Voice Cloning - Clone a custom voice
Voice Design - Generate a voice based on a description

Quick Examples

Text-to-Speech

text_to_audio(
    text="你好，欢迎使用 MiniMax TTS 服务！",
    voice_id="female-shaonv",
    output_path="./hello.mp3"
)

List Available Voices

voices = list_voices(voice_type="system")
for voice in voices:
    print(f"{voice['voice_id']}: {voice['name']}")

Voice Cloning

voice_clone(
    voice_id="my-custom-voice",
    audio_file="./sample.mp3",
    voice_name="我的声音"
)

Voice Design

voice_design(
    prompt="一个温柔的年轻女性声音，带有轻微的南方口音",
    preview_text="你好，这是我的声音"
)

Supported Models

Model	Description
speech-02-hd	HD version, best audio quality
speech-02-turbo	Fast version, low latency
speech-01-hd	Legacy HD
speech-01-turbo	Legacy Fast
speech-2.6-hd	Version 2.6 HD
speech-2.6-turbo	Version 2.6 Fast

Common Voice IDs

System Preset Voices

female-shaonv - Young girl voice
female-yujie - Mature woman voice
female-chengshu - Mature female voice
male-qingnian - Young male voice
male-chengshu - Mature male voice

For more voices, use list_voices() to query.