skills.rest
vellum-ai
vellum-aiOFFICIAL
Explore all skills by this author →
Website𝕏@vellum_aiGitHub

deepgram-voice

Select and configure Deepgram Aura TTS voices for assistant speech output.

1.2k|166|Updated Feb 7, 2026
Search Tags:#text-to-speech#aura#voice selection#deepgram#voice assistant#tts configuration
One-click install
npx skills add https://github.com/vellum-ai/vellum-assistant --skill deepgram-voice
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deepgram-voice
Source: https://github.com/vellum-ai/vellum-assistant/tree/main/skills/deepgram-voice
Command: npx skills add https://github.com/vellum-ai/vellum-assistant --skill deepgram-voice

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve?

Choosing and changing a text-to-speech voice is error-prone because the voice setting lives under whichever TTS provider is active, and writing the wrong config key silently does nothing. This Skill routes voice changes to the correct provider and exposes the full Deepgram Aura voice catalog.

Core Features & Use Cases

  • Curated Voice Selection: Pick from Aura-2 voices (Thalia, Luna, Athena, Zeus, Draco, and more) with style and accent descriptions, then apply via the voice_config_update tool.
  • Provider-Aware Routing: Detects whether the active provider is managed vellum, BYO deepgram, or another provider like ElevenLabs, and writes the voice to the correct config key.
  • Full Catalog Browsing: With a Deepgram API key, query the models endpoint to search voices by tag or accent, preview audio samples, and synthesize custom preview clips.
  • Use Case: A user says their assistant sounds too formal. You check the active provider, suggest aura-2-luna-en for a friendlier tone, apply it with voice_config_update, and the change hot-applies on the next spoken turn.

Quick Start

Change my assistant's voice to a warm British voice using the deepgram-voice skill.

Frequently Asked Questions about deepgram-voice

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I change my assistant's Deepgram TTS voice?▼

Use the voice_config_update tool with setting tts_voice_id and an Aura model id like aura-2-luna-en. The tool writes to the active provider's voice field and pushes the change to the macOS app, applying on the next spoken turn.

What Deepgram Aura voices are available?▼

Curated Aura-2 voices include Thalia, Luna, Athena, and Hera (female) plus Zeus, Orion, Arcas, Apollo, Draco, and Hyperion (male), spanning American, British, and Australian accents. With a Deepgram API key you can query the models endpoint for the full catalog.

Why did my voice change not take effect?▼

The most common cause is writing services.tts.providers.deepgram.model while the active provider is vellum or another provider, since that field is ignored. Check the active provider with assistant config get services.tts.provider and use voice_config_update, which routes to the correct key.

Do I need a Deepgram API key to change voices?▼

No key is needed on the managed vellum provider, which synthesizes Deepgram voices through the platform. A key is only required for the BYO deepgram provider or for browsing the full voice catalog beyond the curated list.

Can I use Deepgram voices with ElevenLabs or other TTS providers?▼

Aura model ids only work with the deepgram or managed vellum providers. To use a Deepgram voice from another BYO provider, switch providers with voice_config_update setting tts_provider to vellum or deepgram first.

Can I adjust Deepgram voice speed or stability?▼

Deepgram Aura has no speed, stability, or similarity parameters; expressiveness is built into each voice. The only tunable is the output audio format (mp3, wav, or opus) on the BYO deepgram provider, so switch voices to change the sound.

Related Skills

Explore More Personal & Entrepreneur Skills→
framerslabframerslab

streaming-stt-deepgram

Transcribe live speech via Deepgram's WebSocket API with speaker diarization.

Official
Advanced
arthtyagiarthtyagi
1

text-to-speech

Generate spoken audio from text using ElevenLabs models and voice settings.

Community
Intermediate
vellum-aivellum-ai
1.0k

elevenlabs-voice

Select and tune ElevenLabs TTS voices with speed, stability, and similarityBoost controls.

Official
Intermediate
malue-aimalue-ai
36

daily-briefing

Generate a daily briefing from weather, calendar, tasks, and news.

Official
Intermediate