mm-easy-voice（mm 简单的声音）

MiniMax Easy Voice

Simple text-to-speech skill powered by MiniMax Voice API. Converts any text into natural-sounding audio with customizable voice selection.

Usage

[Step 1] Preparation

IMPORTANT: Before using this skill, always verify the environment:

python check_environment.py

If API key is not set:

export MINIMAX_VOICE_API_KEY="your-api-key-here"

Requirements:

• Python 3.8+
• MINIMAX_VOICE_API_KEY environment variable (required)
• FFmpeg (optional, for audio merging/conversion)

[Step 2] Generate speech audio from text

Convert any text to speech in one command:

# Default voice (male-qn-qingse)
python mmvoice.py tts "Hello, this is a test." -o test.mp3

# Chinese text
python mmvoice.py tts "你好，这是一个测试。" -o test_cn.mp3

# Specify a different voice by "-v voice_id"
python mmvoice.py tts "Hello world" -v female-shaonv -o hello.mp3

# Specify output path
python mmvoice.py tts "Your text" -o /path/to/output.mp3

Required parameters:

• text: The text you want to convert to speech
• -o OUTPUT: Output audio file path (required)

Optional parameters:

• -v VOICE_ID: Voice to use (default: male-qn-qingse)

Usage Details

Text length limits:

• Up to 10,000 characters per request
• For longer text, split into multiple requests and merge later

Pause insertion: Use <#x#> where x = pause duration in seconds

• Example: "Hello<#1.5#>world" = 1.5 second pause between words
• Range: 0.01 to 99.99 seconds

Emotion matching: speech-2.8 models automatically match emotions to your text content

Voice Selection

Choose the right voice for your content by consulting the voice catalog:

# List all available voices
python mmvoice.py list-voices

Voice catalog: reference/voice_catalog.md

Contains:

• All available system voices (male, female)
• Voice characteristics and recommended use cases
• How to select the right voice for your content

Advanced Options

Voice Management

List available voices:

python mmvoice.py list-voices

Clone a voice from audio sample:

python mmvoice.py clone audio_file.mp3 --voice-id my-custom-voice

Design a voice from description:

python mmvoice.py design "A warm, gentle female voice" --voice-id designed-voice

Audio Processing

Merge multiple audio files:

python mmvoice.py merge file1.mp3 file2.mp3 file3.mp3 -o combined.mp3

Convert audio format:

python mmvoice.py convert input.wav -o output.mp3 --format mp3

Reference Documents

Open these when needed for more details:

|| Document | When to Use | ||----------|-------------| || reference/voice_catalog.md | Choosing a voice_id | || reference/getting-started.md | Environment setup | || reference/audio-guide.md | Audio processing | || reference/voice-guide.md | Voice cloning and design | || reference/troubleshooting.md | Common issues and solutions |

Troubleshooting

Common issues:

1. API key not set: Run export MINIMAX_VOICE_API_KEY="your-key"
2. FFmpeg missing: Install with brew install ffmpeg (macOS) or sudo apt install ffmpeg (Ubuntu)
3. Voice not found: Use python mmvoice.py list-voices to see available voices

Run environment check:

python check_environment.py

See reference/troubleshooting.md for more solutions.