Skip to main content

​
Patter

Connect any AI agent to phone calls with 10 lines of code. Patter handles telephony, speech-to-text, and text-to-speech so you can focus on your agent. Patter is an open-source SDK that gives your AI agent a voice and a phone number. Point it at any function that returns a string and Patter handles the rest — telephony, speech-to-text, text-to-speech, and real-time audio streaming. Works with Python and TypeScript, supports Twilio and Telnyx, and runs in both local and self-hosted modes.

Quick Start

Connect your first AI agent to a phone call in under 5 minutes.

Core Concepts

Understand agents, voice modes, and the audio pipeline.

Python SDK

Full documentation for the Python SDK.

TypeScript SDK

Full documentation for the TypeScript SDK.

​
How it works

Phone Call → Telephony (Twilio/Telnyx) → WebSocket → Patter → STT → Your Agent → TTS → Caller
Your agent receives transcribed speech as text and returns a text response. Patter handles everything else — audio streaming, transcoding, provider APIs, and real-time voice synthesis.

​
Key Features

  • 10 lines of code — Write an async handler. Patter does the rest.
  • Python & TypeScript — Same API surface in both languages. Pick your stack.
  • Multiple voice modes — OpenAI Realtime, ElevenLabs ConvAI, or custom Pipeline (STT + TTS).
  • Twilio & Telnyx — Both telephony providers supported. Bring your own credentials.
  • Call transfer & recording — Built-in system tools for transferring calls and enabling recording.
  • Guardrails — Block terms, run custom checks, or replace responses before they reach the caller.
  • Built-in tunnel — Cloudflare tunnel support, no ngrok required.
  • Real-time dashboard — Live metrics, latency tracking, and cost monitoring.