akillness livekit-agents
LiveKit Agents is a leading voice agent in the 2026 AI agent engineering landscape (11.2k★).
When to use this skill
- Realtime voice AI agents over LiveKit WebRTC rooms
- AgentServer + AgentSession composing VAD/STT/LLM/TTS or realtime speech models
- @function_tool tools callable mid-conversation
- Semantic turn detection model; interruption handling
- Telephony (SIP) integration; avatar integrations
- Deploy to LiveKit Cloud or self-host; Node.js SDK sibling; official Agent Skill (npx skills add livekit/agent-skills)
- Ecosystem overlap: None — voice/realtime is an uncovered category
Installation
pip install "livekit-agents[openai,deepgram,cartesia]"
Core API
from livekit.agents import Agent, AgentServer, AgentSession, JobContext, cli, function_tool, inference
server = AgentServer()
@server.rtc_session()
async def entrypoint(ctx: JobContext):
session = AgentSession(vad=inference.VAD(), stt=..., llm=..., tts=...)
await session.start(agent=Agent(instructions="You are a helpful assistant"), room=ctx.room)
if __name__ == "__main__":
cli.run_app(server)
Instructions
Step 1: Initialize the environment
Verify you have Python installed and set up properly. Use the correct package manager (pip, npm, pnpm, uv) as specified in the installation section.
Step 2: Establish the agent flow
Architecture pattern: Worker/job model: AgentServer registers entrypoints, jobs join rooms; pipeline nodes swappable (any STT/LLM/TTS combo or realtime API); LiveKit Infer provides hosted inference
Step 3: Implement tool integrations and policies
Integrate necessary tools, environment keys, and safety gates according to your architectural needs. Position in the ecosystem: Infrastructure-grade voice-agent lane (powers ChatGPT voice); alternatives: pipecat (pipeline flexibility), vapi/retell (managed SaaS)
Examples
Example 1: Minimal Agent Run
from livekit.agents import Agent, AgentServer, AgentSession, JobContext, cli, function_tool, inference
server = AgentServer()
@server.rtc_session()
async def entrypoint(ctx: JobContext):
session = AgentSession(vad=inference.VAD(), stt=..., llm=..., tts=...)
await session.start(agent=Agent(instructions="You are a helpful assistant"), room=ctx.room)
if __name__ == "__main__":
cli.run_app(server)
Best practices
- Follow the primary installation commands for LiveKit Agents to ensure dependencies are resolved correctly.
- Configure model providers and API keys securely before starting the runtime.
- Use appropriate sandboxing or permission isolation if executing AI-generated code.