The Open-Source Voice AI Stack
Build and deploy production-ready, multi-tenant AI voice assistants in minutes, not months.
Build and deploy production-ready, multi-tenant AI voice assistants in minutes, not months.
"I was a vapi.ai power user for 2.5 years. The UI never worked properly, latency was terrible (4-5 seconds), and their recent pricing changes would cost my startup $800/month. With Burki, I get sub-second response times and complete control over my costs."
— Real user experience that inspired Burki
Burki provides the entire stack you need to build, deploy, and manage voice agents at scale.
Manage multiple organizations and assistants from a single deployment, with complete data isolation.
An integrated dashboard provides instant insights into call volume, performance, and costs.
Built-in RNNoise denoising and optimized streaming deliver 0.8-1.2s response times. No more waiting 4-5 seconds for responses.
Integrate with your favorite LLM, TTS, and STT providers. Own your data and infrastructure—no vendor lock-in or hidden costs.
A complete REST API and webhooks allow for deep integration and process automation.
Get a full production-ready instance running in minutes with our Docker-based deployment.
From cloning the repo to a running instance in minutes. No weeks of API integration like with vapi.ai.
Clone the repo and add your provider API keys to the `.env` file.
Run the application and use the web dashboard to create your first assistant.
Get started with the world's most powerful open-source voice AI platform today.