elevenlabs-voice_skill

This skill converts text to natural speech and transcribes audio using ElevenLabs, enabling multilingual TTS and accurate STT for messages.

Python

2.6k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 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 elevenlabs-voice

_meta.json357 B
SKILL.md4.7 KB

Overview

This skill provides a complete voice solution using ElevenLabs AI for Text-to-Speech (TTS) and Speech-to-Text (STT). It delivers high-quality, configurable AI voices and accurate transcription across many languages. Use it to generate natural-sounding audio, transcribe voice messages, or build voice-enabled workflows.

How this skill works

The skill calls ElevenLabs APIs to perform TTS with selectable voices, stability, and similarity settings, producing MP3 or other supported audio files. For STT it uses ElevenLabs Scribe to transcribe uploads, with optional language hints and speaker diarization. It includes helper client classes and command-line scripts to integrate the functionality into apps or bots.

When to use it

Convert text responses into natural-sounding audio for bots, apps, or notifications.
Transcribe user voice messages, interviews, or meetings into text for search and analysis.
Build multilingual voice features using models optimized for non-English languages.
Create automated voicemail, accessibility features, or audio previews from text.
Prototype voice interactions in chat platforms like Telegram or integrate into workflows.

Best practices

Set and secure your ELEVENLABS_API_KEY in environment variables or a .env file.
Choose voice and model based on target language: use multilingual models for non-English audio.
Tune stability and similarity_boost to balance expressiveness and consistency for your use case.
Respect file size limits (<=100 MB) and supported formats when uploading audio for transcription.
Cache common generated audio and transcriptions to reduce API calls and costs.

Example use cases

Bot replies: transcribe a Telegram voice note, generate a spoken reply, and send it back as a voice message.
Content production: convert blog posts or articles into narrated audio files with selectable voices.
Customer support: transcribe support calls and create concise text summaries for agents.
Multilingual apps: produce localized audio using the multilingual model for non-English users.
Accessibility: generate audio descriptions or read-aloud content for visually impaired users.

FAQ

Supported formats include mp3, mp4, mpeg, mpga, m4a, wav, and webm. Max file size is 100 MB.

How do I choose the right voice and model?

Pick a voice by its voice_id for the desired tone. Use eleven_multilingual_v2 for non-English audio, eleven_turbo_v2_5 for fast high-quality output, and eleven_monolingual_v1 for English-only scenarios.