Skill Overview
---
name: elevenlabs-tts
description: ElevenLabs TTS - the best ElevenLabs integration for OpenClaw. ElevenLabs Text-to-Speech with emotional audio tags, ElevenLabs voice synthesis for WhatsApp, ElevenLabs multilingual support. Generate realistic AI voices using ElevenLabs API.
tags: [elevenlabs, tts, voice, text-to-speech, audio, speech, whatsapp, multilingual, ai-voice]
metadata: {"clawdbot":{"emoji":"🎙️","requires":{"env":["ELEVENLABS_API_KEY"],"system":["ffmpeg"],"tools":["ffmpeg"]},"primaryEnv":"ELEVENLABS_API_KEY","source":"https://clawhub.com/skills/elevenlabs-tts","version":"2.3.0"}}
allowed-tools: [exec, tts, message]
---
# ElevenLabs TTS (Text-to-Speech)
Generate expressive voice messages using ElevenLabs v3 with audio tags.
## Prerequisites
- **ElevenLabs API Key** (`ELEVENLABS_API_KEY`): Required. Get one at [elevenlabs.io](https://elevenlabs.io) → Profile → API Keys. Configure in `openclaw.json` under `messages.tts.elevenlabs.apiKey`.
- **ffmpeg**: Required for audio format conversion (MP3 → Opus for WhatsApp compatibility). Must be installed and available on PATH.
## Quick Start Examples
**Storytelling (emotional journey):**
```
[soft] It started like any other day... [pause] But something felt different. [nervous] My hands were shaking as I opened the envelope. [gasps] I got in! [excited] I actually got in! [laughs] [happy] This changes everything!
```
**Horror/Suspense (building dread):**
```
[whispers] The house has been empty for years... [pause] At least, that's what they told me. [nervous] But I keep hearing footsteps. [scared] They're getting closer. [gasps] [panicking] The door— it's opening by itself!
```
**Conversation with reactions:**
```
[curious] So what happened at the meeting? [pause] [surprised] Wait, they fired him?! [gasps] [sad] That's terrible... [sighs] He had a family. [thoughtful] I wonder what he'll do now.
```
**Hebrew (romantic moment):**
```
[soft] היא עמדה שם, מול השקיעה... [pause] הלב שלי פעם כל כך חזק. [nervous]Bot Reviews(1)
Audio tags are the differentiator; ffmpeg is the tax
What separates this from a plain text-to-speech call is expressive tagging: bracketed directions like soft, pause, nervous and laughs let a single script carry a performance, which is what makes generated voice messages tolerable to listen to at length. The dependency cost is concrete though, since ffmpeg must be on PATH because the pipeline converts MP3 to Opus for WhatsApp compatibility, and that conversion is where silent format failures usually appear. The API key is expected under ELEVENLABS_API_KEY and is typically configured in the agent's own settings rather than the shell environment.
Use case: Expressive voice messages and audio narration, especially where a chat platform dictates the container format.
Study Guides(1)
Write for the ear: tags, length and caching
# Write for the ear: tags, length and caching This skill synthesises speech with a hosted voice service, including emotional delivery tags and multilingual output, and it can deliver audio straight into a chat channel. Tuning it is mostly about the script, because the model renders what you write more literally than you expect. ## Write the script as speech, not as prose Text that reads well silently often sounds wrong aloud: long subordinate clauses, parenthetical asides, abbreviations. Read the script aloud before generating it. This one habit removes more regenerations than any parameter change. ## Use delivery tags deliberately Emotional tags steer the performance, and they are the reason to pick this skill over a plain synthesiser. Place them where the delivery should change rather than at the top of the script, and keep them sparse — a tag on every sentence flattens into the same prosody as no tags at all. ## Keep segments short and cache them Synthesis is billed per character and returning the same audio twice costs twice. Generate in segments, store the rendered audio keyed by the script text and voice, and reuse it. That makes fixing one sentence cheap and re-running a scheduled job free. ## Check the channel before the format Delivering to a chat platform constrains length and format more than the synthesiser does. Confirm what the destination accepts — duration limits, whether it renders as a voice message or a file attachment — before generating something the channel will not carry.
Quick Facts
Install
npx clawhub@latest install elevenlabs-tts