Self-hostable open-source voice AI platform for building, testing and deploying phone and browser voice agents, with a visual workflow builder and MCP support.
- Stars
- 1,078
- Forks
- 126
- Open issues
- 11
- Last commit
- 19 days ago
What is it?
Patter is an SDK for building AI voice agents that answer and place real phone calls. You write the agent logic, and Patter handles the agent loop, the language model, speech-to-text, text-to-speech, real-time voice, audio processing, and the telephony carrier connection.
Each layer is pluggable. Developers can choose providers for the LLM, STT, TTS, all-in-one realtime engine, and carrier (Twilio, Telnyx, or Plivo), and switch between them with a single line. Realtime, Pipeline, and Hybrid modes are supported, and the Python and TypeScript APIs are designed to match.
For local development, the SDK can start a built-in tunnel and dashboard, or simulate a call from the terminal without a phone. Tools, call transfer, guardrails, and an OpenTelemetry trace of each call are included, and telemetry can be disabled.
Key features
- 01Gives an AI agent a phone number with Python or TypeScript code.
- 02Lets you swap LLM, STT, TTS, realtime engine, and carrier providers.
- 03Supports Twilio, Telnyx, and Plivo for telephony.
- 04Includes tools, call transfer, guardrails, and LLM fallback chain.
- 05Provides a local tunnel, dashboard, and terminal call simulation.
Use cases
- Building an inbound phone assistant, such as a restaurant or clinic booking line.
- Placing outbound calls with answering machine detection and voicemail drop.
- Connecting a voice agent to a CRM through webhook tools.
- Prototyping voice agents locally before deploying with a static webhook URL.
Quick start
- 1
Install the Python SDK from PyPI
pip install getpatter - 2
Create a phone agent with Twilio and OpenAI Realtime, then serve it with a tunnel
- 3
Or install the TypeScript SDK from npm
npm install getpatter
Things to know
- Carrier and AI provider credentials must be supplied via environment variables, so you need accounts with those services.
- Local development relies on a tunnel, and production needs a static webhook URL or a separate tunnel service.
In collections
Topics
FAQ
- What is Patter?
- Open-source Python and TypeScript SDK that gives an AI voice agent a phone number, handling LLM, speech, and carrier layers. Patter is an SDK for building AI voice agents that answer and place real phone calls. You write the agent logic, and Patter handles the agent loop, the language model, speech-to-text, text-to-speech, real-time voice, audio processing, and the telephony carrier connection.
- Is Patter free to use? What license does it have?
- Patter is open source under the MIT license. Check the license terms before using it in commercial products.
- Is Patter still maintained?
- Actively maintained. The last commit to Patter was on Oct 3, 2026.
- What can I use Patter for?
- Building an inbound phone assistant, such as a restaurant or clinic booking line. Placing outbound calls with answering machine detection and voicemail drop. Connecting a voice agent to a CRM through webhook tools. Prototyping voice agents locally before deploying with a static webhook URL.
- How do I install Patter?
- Install the Python SDK from PyPI: pip install getpatter Create a phone agent with Twilio and OpenAI Realtime, then serve it with a tunnel Or install the TypeScript SDK from npm: npm install getpatter
- Where can I get Patter?
- The source code and installation instructions are on GitHub: https://github.com/PatterAI/Patter
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A week-by-week course for building an AI voice call center with FastRTC, Superlinked, Twilio, Opik and Runpod.
Open-source API and voice pipeline that gives AI agents a real phone number, SMS, and email inbox, with a bring-your-own LLM option.
Self-hosted gateway that connects phone lines and chat channels to AI agents, using your own API keys.
Open-source Python framework for building real-time voice and multimodal AI agents, with pluggable speech, LLM, and transport services.
A Python framework for building realtime voice and video AI agents that run as server-side participants in LiveKit rooms.