A guide exploring why conversational voice agents often drop or fail to respond to short user utterances like 'yes' or 'ok', and how to debug the underlying pipeline.

LiveKit has transitioned to per-second billing with a ten-second minimum increment for WebRTC connections, SIP calls, agent sessions, and recordings.

Protect voice agent recordings and transcripts with PII Redaction, available at no extra cost with LiveKit Agent Observability.

SIP and RTP get media into a voice-AI pipeline, but the control plane — what the agent is told and decides to do next — is where most production bugs actually live.

The annual ClueCon conference founded by the creators of FreeSWITCH has always been where the real-time communications community gathers to share what’s working, what’s breaking, and what’s next. This year’s edition, August 10-14 at The Midland Hotel in Chicago, made one thing un

Robots, policies, operators, and recorders as independent participants in one realtime session: five multi-participant architectures we have helped customers build on LiveKit.

The moq.dev project has moved from a side effort to a funded, full-time push on Media over QUIC — kixelated is now building it as a business, not just a spec implementation.

GPT-Live-1 makes full duplex voice AI the new baseline, and it is going to be hard to beat with today's turn taking architecture.

Expressive mode is a single flag that lets the LLM shape how your agent sounds, not just what it says. Tell it good news or bad news and hear the delivery change.

Human speech is often unstructured and messy. To keep up with real conversation, Voice AI bots must parse partial sentences, filler words, and interruptions, all without the visual cues a human listener relies on. On top of that linguistic challenge, real-world latency budgets an

In the space of thirteen months, Google took real-time speech translation from an English-and-Spanish demo at I/O 2025 to Gemini 3.5 Live Translate, a 70-language streaming model announced in June 2026 and now expanding into Google Meet through private preview, on top of the five

We built a self-hosted video calling system for a healthcare platform client, replacing the usual “meeting link” pattern with a real phone-call flow: clinician dials, patient’s phone rings, patient answers. Highlights: Read the full WebRTC.ventures Story of Success: Self-Hosted J

STUNner is a Kubernetes-native TURN server, TURN being the relay protocol that gives WebRTC clients a public endpoint to send media through. It builds on the Gateway API to terminate TURN at a single load balancer endpoint and relay media to the pod behind it, so media-server […]

Twilio SIGNAL 2026 was not a new channel. It was a repositioning: the Conversation Layer, what it means for WebRTC, and the lock-in nobody named on stage.

How to detect if your agent is slow to join a room, and fix it.

Swapping WebRTC stacks doesn't fix call quality you can't measure — the argument for treating observability, not the transport layer, as the real lever on perceived quality.

Voice AI tooling broadly falls into two categories: managed platforms that get you a production agent fast, and custom-built frameworks that give you full ownership of your voice pipeline. Within each category, the specific tools differ in ways that are worth understanding before

WebRTC.ventures announced upcoming ClueCon presentations focusing on real-time voice AI telephony architecture and utilizing eBPF to optimize TURN server performance.

An overview of how to build a compliance architecture for production telephony AI agents using Retell AI as a reference implementation.

WebRTC Live #115 explores how WebRTC, voice AI, and interactive avatars were combined to build a real-time simulation platform for training frontline mental health providers.

Every vendor sees only the deals that reach them. The 2026 Video API Report maps the market no insider can see and no buyer has time to build.

A playful but genuinely useful explainer that maps MoQ's design goals onto a Pixar analogy — good for building intuition on why MoQ trades strict ordering for low latency.

I set out to get one tool seen and ended up mapping the whole WebRTC tools ecosystem, scored by how AI actually sees it.

tl;dr: Opus encoding just got 20% faster on desktop Chromium by turning on AVX2 vectorization in libopus. Vectorization means one instruction doing the same arithmetic on a whole vector of values at once instead of one at a time. Codecs spend their hot loops running identical mat

LiveKit's Turn Detector v1 listens to speech directly instead of waiting on a transcript, fusing semantic and acoustic signals for end-of-turn detection across 14 languages.

Most buyers pick a Video API vendor on features. The pricing model is what they get stuck with.

Well, that title is clickbait for Hacker News. The honest title would have been “using AI to tackle long-standing WebRTC issues.” Posts on webrtcHacks are rare these days, and in German we would say there are Unkenrufe that WebRTC is dead. It is not. It is the infrastructure for

Chrome moves to a 2 week release cycle. Where are you with your WebRTC app

Tsahi Levent-Levi examines the current state of Media over QUIC (MoQ), arguing that the emerging standard still lacks a clear catalyst or killer use case to drive widespread industry adoption.

There are ways to do voice AI without being traumatized by WebRTC.

An analysis of why traditional backend infrastructure metrics fail to capture real-time media degradation and how to implement client-side WebRTC monitoring.

Using MoQ for REAL-TIME BIDIRECTIONAL STREAMING. Gameboy games!

An analysis of how AI-driven surges in memory prices are impacting infrastructure economics and architectural decisions across WebRTC and real-time Voice AI deployments.

First class support for MoQ in OBS... via a fork. Go forth and integrate libmoq into all of your favorite tools still using C for some reason.

Do you need to use MoQ? What are we even doing with our lives?

It turns out there's actually a bunch of people behind MoQ not named `kixelated`

Maybe join us in Uruguay this year? We'll be working on MoQ and stuff.

A level-headed breakdown of when WebRTC's sub-second interactivity still wins and when MoQ's CDN-friendly, QUIC-native fan-out is the better fit — not a 'MoQ replaces WebRTC' piece.

It is time for another edition of “Is everyone switching to…“. Cloudflare recently published a blog post about Media over Quic (MoQ) which made a number of statements about WebRTC that require some “clarification”. Let us start with that and look at MoQ and WebTransport after tha

Earlier this month, OpenAI released the GA version of its realtime API. This includes many capabilities that the Beta didn’t have, including video support. I started out doing an update to the The Unofficial Guide to OpenAI’s Realtime WebRTC API I made for the Beta release last N

An introduction to hang.live, an open-source real-time video conferencing application built entirely on Media over QUIC (MoQ) without relying on WebRTC.

Cloudflare has launched a technical preview of its Media over QUIC (MoQ) CDN, offering global edge infrastructure for low-latency media delivery.

A practical guide navigating the complexities and common pitfalls of configuring TLS 1.3 for QUIC and WebTransport deployments.

An essay on how software engineers can navigate generative AI tools by leveraging baseline outputs while focusing personal effort on being an outlier in creativity and problem-solving.

Audio jitter buffers are required 101 introductory material for understanding VoIP. libWebRTC’s audio jitter buffer implementation – the one in Chromium – is known as NetEQ. NetEQ is anything but basic. This is good from a user perspective since real-life networks conditions are

OpenAI is utilizing WebRTC for its Realtime API! Even better, webrtcHacks friend and Pion founder Sean DuBois helped to develop it and agreed to a Q&A about the implementation. It is not often a massive WebRTC use case like this emerges so rapidly. In addition, Sean was extremely

An empirical analysis measuring the end-to-end voice response latency of OpenAI's WebRTC-based Realtime API compared to chatgpt.com.

A technical deep dive into OpenAI's Realtime API architecture and its implementation of WebRTC for low-latency bidirectional voice interactions.

A quantitative analysis of 2024 GitHub event data tracking open-source WebRTC repository activity, popular frameworks, and API adoption trends over the past decade.

An architectural breakdown of the layered protocol stack powering Media over QUIC (MoQ), tracing media delivery down through WebTransport, QUIC, and lower layers.

An overview of overlooked and lesser-known features in the QUIC transport protocol that make it compelling for next-generation media delivery.