mm-easy-voice_skill

This skill converts text to natural speech using MiniMax Voice API with customizable voices for quick audio generation.

Python

2.6k

GitHub Stars

5

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill mm-easy-voice

_meta.json636 B
check_environment.py2.3 KB
mmvoice.py17.9 KB
requirements.txt96 B
SKILL.md3.5 KB

Overview

This skill provides a simple text-to-speech interface using the MiniMax Voice API to convert text into natural-sounding audio files. It supports selectable voices, pause insertion, emotion-aware models, and basic audio processing like merging and format conversion. Setup is minimal: Python 3.8+, an environment variable for the API key, and optional FFmpeg for advanced audio handling. The command-line tool produces mp3/wav outputs and includes voice management utilities.

How this skill works

The tool sends text payloads to the MiniMax Voice API and receives synthesized audio, saving it to the specified output path. You can choose a voice_id or use the default; speech-2.8 models will match emotion automatically. Long texts are split into multiple requests and can be merged locally. Additional commands let you list voices, clone or design voices from samples, and perform basic audio conversions and merges with FFmpeg when available.

When to use it

Generate narration for videos, podcasts, or e-learning content.
Produce quick voice previews or prototypes for voice UX decisions.
Automate spoken notifications, IVR prompts, or accessibility audio.
Create localized or bilingual audio by switching voice/text language.
Merge multiple generated clips into a single track for long-form content.

Best practices

Set MINIMAX_VOICE_API_KEY in your environment before running commands.
Keep requests under 10,000 characters; split and merge longer texts to avoid errors.
Choose voice_id from the voice catalog to match tone and audience; use list-voices to explore options.
Use the <#x#> pause tag for natural pacing and test pause durations in small samples.
Install FFmpeg if you need merging or format conversions for higher-quality output.

Example use cases

Turn article text into an mp3 narration for a blog post using a warm female voice.
Generate multilingual prompts for an app—Chinese lines with a suitable Chinese voice and English lines separately.
Clone a client’s brand voice from a recorded sample for consistent automated messages.
Create an audiobook by splitting a long manuscript into chunks, synthesizing each, then merging into a single file.
Produce short voice assets for notifications and export them in multiple formats (mp3/wav).

FAQ

You need Python 3.8+, the MINIMAX_VOICE_API_KEY environment variable set, and optionally FFmpeg for merging or format conversion.

How do I control pauses in speech?

Insert <#x#> where x is the pause in seconds (0.01 to 99.99), for example "Hello<#1.5#>world" adds a 1.5s pause.

What if my text is longer than the limit?

Split the text into multiple requests (each <=10,000 characters), synthesize each part, then merge the resulting audio files.

How do I find and pick voices?

Run python mmvoice.py list-voices to view available voices and consult the voice catalog for characteristics and recommended use cases.