Voice Agents Cheat Sheet

This content was generated with the assistance of AI and should be verified against the official documentation of each tool and framework before being relied on in production.

This section’s bibliography lists the reference material consulted while preparing these pages.

A single-page index of quick back-links to every page in this sub-section, grouped the same way as the navigation. It was written on 2026-10-01 against LangChain 1.4.3 with FastAPI 0.142.2, Spring AI 2.0.1 on Spring Boot 4.1.x, Pipecat 1.12.0, LiveKit Agents 1.8.3, OpenAI gpt-live-1 and the Realtime API (gpt-realtime-2.1), gemini-3.8-live, the ElevenLabs Python SDK 2.70.0, Twilio ConversationRelay and Media Streams, and Asterisk 23.5, 22.11 and 20.21; the full list is the version baseline on the landing page.

The downloadable PDF is a separate, single A4 page meant for printing. It states the version baseline and covers:

  • the cascaded, speech-to-speech and hybrid architectures, and the latency budget formula;

  • the turn-taking checklist: VAD, endpointing, barge-in and echo;

  • speech-to-text and text-to-speech providers and local models, with their licences;

  • audio formats and resampling for telephone, browser and model paths;

  • the realtime API transports of OpenAI, Gemini Live and Azure Voice Live;

  • the ElevenLabs custom LLM setup;

  • Pipecat and LiveKit Agents skeletons;

  • the Twilio ConversationRelay and Media Streams message tables;

  • the Asterisk ARI External Media, AudioSocket and chan_websocket transports;

  • the Spring AI audio interfaces (TranscriptionModel, TextToSpeechModel);

  • the production checklist.

Overview — Voice Agents.

Foundations — Voice Architectures and Turn-Taking and Interruptions.

Speech models — Speech to Text and Text to Speech.

Hosted platforms — Realtime APIs and ElevenLabs Agents.

Frameworks — Pipecat, LiveKit Agents and Building a Voice Pipeline by Hand.

Web delivery — Voice in the Browser.

Telephony — Telephony with Twilio and Telephony with Asterisk and SIP.

Operations — Production Voice Agents.

The one formula the whole sheet builds on — the time from the end of the caller’s sentence to the first audible word of a cascaded pipeline (explained on Voice Architectures):

Source: LiveKit Agents documentation and Pipecat documentation for the stage metrics, and Spring AI — Text-To-Speech (TTS) API.

Download the Voice Agents Cheat Sheet (PDF)

References

The PDF summarises the pages above; each source is cited in full on the page that covers it, and in the bibliography.