Pipecat Cloud Agent Images

This repository contains source code for the official Pipecat Cloud agent images and starter templates.

Repository Structure

• pipecat-base: Source code for the dailyco/pipecat-base Docker image, which serves as the foundation for Pipecat Cloud agents.
• pipecat-starters: Ready-to-use agent templates for various use cases.

Base Image

The base image provides the runtime environment and interface required to run agents on Pipecat Cloud. It handles:

• Starting agent processes in response to API calls
• Session management
• Platform integration
• Logging and monitoring

The official base image is available on Docker Hub: dailyco/pipecat-base

Starter Templates

Starter templates offer pre-built configurations for common agent types:

• voice: Voice conversation agent with STT, LLM and TTS
• twilio: Telephony agent that works with Twilio
• pstn_sip: Dial-in or out to your agent using Daily's telephony offerings
• natural conversation: Text chat agent focused on natural dialogue
• openai realtime: Agent using OpenAI's streaming capabilities
• gemini multimodal live: Multimodal agent using Google's Gemini models
• vision: Computer vision agent that can analyze images

Each starter includes a functioning implementation and Dockerfile. They serve as a starting point for building an agent tailored to your use case.

Documentation

For detailed instructions on using these images and deploying agents:

• Agent Images Guide
• Quickstart Guide