Configure audio streams
Write WebSocket code to stream speech-to-text with proper keepalive frames.
Configure speech APIs and deploy voice integrations instantly with specialized AI coding agent skills.
Official Agent Skills published by Deepgram
Equip your AI coding agent with the verified instructions needed to deploy real-time audio transcription and build low-latency voice apps using Deepgram.
Write WebSocket code to stream speech-to-text with proper keepalive frames.
Connect transcription APIs to Twilio and LiveKit communication networks.
Implement low-latency text-to-speech outputs and turn-taking models.
Official Agent Skills published by Deepgram.
Provides direct access to the Deepgram API reference to help build speech-to-text, text-to-speech, voice agents, and audio intelligence features.
Finds and retrieves official Deepgram documentation, API references, and migration guides for speech-to-text, text-to-speech, and voice agents.
Provides runnable, end-to-end integration examples for Deepgram APIs with third-party platforms, telephony, and web frameworks.
Add Deepgram Skills to agents that support the Agent Skills format, including Claude, Claude Code, Codex, and OpenCode. Install manually from source, or use MCP Market Hub to sync and share them with your team.
Add official skills once, keep them current across your agents, and share the same skill set with your team.
Prefer manual control? Download the skills from source, then follow the install guide for your agent.
deepgram/skillsReal workflows powered by Deepgram skills.
Configure real-time transcription
Build streaming audio connections using Nova and Flux models for live transcription.
Set up voice agents
Create conversational assistants with low latency turn-taking using telephony integrations.
Find runnable code recipes
Retrieve verified SDK snippets under 50 lines in Python, JavaScript, and Go.
Verified, first-party skills from other brands you know.