LiveKit

What it does

LiveKit takes realtime audio and video from users who connect through an app, a browser or a phone call, and passes that speech to an agent you write in Python or Node.js. Your agent runs your business logic, calls language, speech-to-text and text-to-speech models, and streams a spoken or video response back to the user. The open source Agents framework and media server can run on your own infrastructure, while LiveKit Cloud hosts and scales agent deployments.

Developers typically start with the Voice AI quickstart, which builds a simple voice agent with Python or Node.js, then move to starter apps for web or mobile and connect telephony to take or place phone calls. Built-in features include automatic turn detection and interruption handling, an inference gateway that gives one API key access to LLM, STT and TTS models, and session observability with transcripts, traces and logs.

LiveKit suits developers and engineering teams building voice AI, video AI or robotics applications who want code-level control over agent logic and model providers. The Build plan is free at $0/month and includes 1,000 agent session minutes per month, with usage beyond allowances metered per minute. Model costs are billed separately per model, and self-hosting requires your own infrastructure and compute. SSO, a HIPAA BAA and on-premise deployment are available through sales.

What it can do

• Build voice AI agents in Python or Node.js with turn detection and interruption handling
• Access LLM, STT, TTS and turn handling models through a single LiveKit Inference API key
• Deploy and autoscale agents on LiveKit Cloud with instant rollback and deployment metrics
• Connect agents to phone calls through US local or toll-free numbers and SIP integrations
• Record and export session audio, transcripts, traces and logs for testing and observability
• Run the open source LiveKit Agents framework and media server locally or on your own infrastructure

Works with

• Twilio
• WhatsApp Business
• AWS PrivateLink
• Azure PrivateLink
• Amazon S3
• Krisp
• ai-coustics

Example workflows

Build a voice support agent

Input: a developer's Python or Node.js agent code and a system prompt. Steps: follow the Voice AI quickstart, choose LLM, STT and TTS models through LiveKit Inference, deploy to LiveKit Cloud. Output: a deployed agent that answers spoken user questions in a web or mobile starter app.

Connect an agent to a phone number

Input: a deployed voice agent and a US local phone number. Steps: rent a number in the dashboard, attach it to the agent through the telephony integration, test inbound calls. Output: an agent that takes phone calls from customers, with inbound minutes metered against the plan.

Before you choose

• Costs are per minute for agent sessions, telephony, models and observability, so estimate with the pricing calculator before scaling.
• Model prices are separate from platform plans and vary by provider and model; check the inference pricing page for current rates.
• Self-hosting the open source framework does not remove the cost of running models or compute yourself.

Listed under

• Coding
• Sandboxes and infrastructure
• Agents
• Developer tools
• Open source
• Realtime
• Telephony
• Video AI

Get started

A free account is available with no credit card required; the Voice AI quickstart is the documented first step.

1. Create a LiveKit Cloud account from the Start building link.
2. Follow the Voice AI quickstart to build and run a simple agent locally with Python or Node.js.
3. Deploy the agent to LiveKit Cloud and optionally add a starter app or telephony number.

Build is $0/month with 1,000 agent session minutes. Ship starts at $50/month and Scale at $500/month. Model costs (LLM, STT, TTS) and telephony are charged per minute on top; the Enterprise plan is custom pricing via sales. Check the pricing page and inference pricing for current rates.

Source: livekit.com/pricing, checked 29 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Build voice, video, and physical AI | LiveKit ↗livekit.com
• Pricing | LiveKit ↗livekit.com

Company website

livekit.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.