openclaw-telnyx-tts

Telnyx TTS speech provider for OpenClaw — 10x cheaper than ElevenLabs, carrier-grade latency.

Use Telnyx AI as your OpenClaw text-to-speech backend. Every /tts command, voice note, and ClawdTalk voice call will use Telnyx TTS — premium NaturalHD voices at a fraction of ElevenLabs' cost.

Why Telnyx TTS?

Quick Start

1. Install

openclaw install npm:openclaw-telnyx-tts

Or manually:

npm install -g openclaw-telnyx-tts

2. Set your API key

export TELNYX_API_KEY="KEY..."

Or in openclaw.json:

{
  "models": {
    "providers": {
      "telnyx": {
        "apiKey": "KEY..."
      }
    }
  }
}

3. Restart the gateway

openclaw gateway restart

That's it. TTS will automatically detect and use Telnyx when TELNYX_API_KEY is set.

How It Works

This plugin registers telnyx as a first-class speech provider in OpenClaw's TTS system. Under the hood:

1. Auto-detection: When TELNYX_API_KEY is set, the plugin is auto-selected with priority 15 (before ElevenLabs at 20).
2. WebSocket TTS: Uses wss://api.telnyx.com/v2/text-to-speech/speech for low-latency streaming synthesis.
3. Zero config: No messages.tts config block needed. Just set the API key and restart.
4. Full contract: Implements synthesize, synthesizeTelephony, listVoices, and directive parsing.

Configuration

Explicit provider selection

Force Telnyx as the TTS provider:

{
  "messages": {
    "tts": {
      "provider": "telnyx"
    }
  }
}

Custom voice

{
  "messages": {
    "tts": {
      "providers": {
        "telnyx": {
          "voice": "Telnyx.NaturalHD.luna"
        }
      }
    }
  }
}

Available voices

Voice directives

Override the voice per-message using directives:

[[voice:Telnyx.NaturalHD.luna]] Hello, this is Luna speaking.

Custom base URL

For proxy setups:

{
  "messages": {
    "tts": {
      "providers": {
        "telnyx": {
          "baseUrl": "wss://your-proxy.example.com/v2/text-to-speech"
        }
      }
    }
  }
}

Auto-Select Priority

When TTS provider is "auto" (the default), OpenClaw tries speech providers in order:

Verify It's Working

After installation and restart, test with:

/tts Hello from Telnyx!

Or check provider status:

/tts-status

Troubleshooting

"Telnyx API key missing"

Set your API key:

export TELNYX_API_KEY="KEY..."

Get one at portal.telnyx.com.

"WebSocket error" or timeout

• Check your network connection
• Ensure wss://api.telnyx.com is reachable
• Try increasing timeout in config

TTS still using ElevenLabs

• Make sure the plugin is installed: openclaw plugins list
• Restart the gateway: openclaw gateway restart
• Remove any explicit "provider": "elevenlabs" in your TTS config

Development

git clone https://github.com/team-telnyx/openclaw-telnyx-tts.git
cd openclaw-telnyx-tts
npm install
npm run build

Testing locally

npm link
openclaw install link:openclaw-telnyx-tts

Architecture

┌──────────────────────────────────────────────────┐
│                OpenClaw Gateway                │
│                                                  │
│  /tts "Hello!"  or  voice-note synthesis         │
│       │                                          │
│       ▼                                          │
│  ┌─────────────────────┐                         │
│  │  Speech Provider    │  auto-select chain:     │
│  │  Manager            │  telnyx → elevenlabs    │
│  └────────┬────────────┘  → microsoft            │
│           │                                      │
│           ▼                                      │
│  ┌─────────────────────────────────────┐         │
│  │  telnyx-tts plugin                  │         │
│  │  (this package)                     │         │
│  │                                     │         │
│  │  buildTelnyxSpeechProvider() →      │         │
│  │    telnyxTts({                      │         │
│  │      voice: "NaturalHD.astra",     │         │
│  │      apiKey: TELNYX_API_KEY         │         │
│  │    }) → mp3 Buffer                  │         │
│  └────────┬────────────────────────────┘         │
│           │                                      │
└───────────┼──────────────────────────────────────┘
            │ WebSocket (wss://)
            ▼
   ┌────────────────────┐
   │  Telnyx TTS API    │
   │  /v2/text-to-      │
   │  speech/speech     │
   │                    │
   │  NaturalHD voices  │
   └────────────────────┘

License

MIT

Credits

Built for the OpenClaw ecosystem by the Telnyx AI FDE team.