Founders @jordan_dearsley and @nikhilro_ sat down with @gustaf at YC to tell the Vapi story. Many pivots, including a Zoom button and AI therapy, led to voice agents. Now a billion calls a year and the goal to help every business build a human interface. piped.video/watch?v=2otGNwNO…

We’re excited to announce that Andy O'Dower, VP and Field CTO at @Twilio, will be speaking at VapiCon 2026! Andy leads the work of turning agentic AI into production communications infrastructure. He partners with Twilio's R&D and go-to-market leadership on the platform's AI direction, and previously ran product for Voice, Video, and Flex, used by millions of developers and generating hundreds of millions in annual revenue. Andy was a Twilio customer before he was a Twilio executive, having built on the platform at venture-backed startups and public companies. He has spent more than 20 years building software platforms and the teams behind them. Join us in San Francisco this November to hear Andy and other industry leaders discuss the future of Voice AI. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

We’re excited to welcome Phonic as a Startup Sponsor of VapiCon 2026! Phonic is the platform for building, evaluating, and deploying voice AI agents that sound like real conversations. Its speech models are designed for human-like emotion, timing, and nuance, with low-latency responses and natural interruption handling. We’re thrilled to have the Phonic AI team joining the Voice AI community at VapiCon. Join us November 11–12 at Fort Mason in San Francisco as the voice AI community connects, builds, and launches what’s next. Learn more and purchase your VapiCon 2026 ticket at vapicon.ai/

We’re excited to welcome @Modulate_ai as a Startup Sponsor of VapiCon 2026! Modulate is a voice intelligence company building AI models and APIs designed to understand real-world conversational audio at scale. Its technology combines speech recognition, acoustic analysis, and conversational context to deliver reliable, explainable, and cost-effective voice intelligence for developers and enterprises. We’re thrilled to have the Modulate AI team joining the Voice AI community at VapiCon. Join us November 11–12 at Fort Mason in San Francisco as the voice AI community connects, builds, and launches what’s next. Learn more and purchase your VapiCon 2026 ticket at vapicon.ai

We're thrilled to announce that @grinich, Michael Grinich, is going to be speaking at VapiCon 2026! Michael is the Founder and CEO of @WorkOS, a developer platform powering enterprise features for leading AI companies—including OpenAI, Anthropic, Perplexity, and Cursor—as well as hundreds of other modern SaaS applications. Join us in San Francisco this November as Michael and other industry leaders discuss the future of Voice AI. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

When we hosted VapiCon last year, voice AI was early. Today it's production-ready, and teams are deploying at scale with results that speak for themselves. VapiCon shows you what's next. Fort Mason, SF · Nov 11–12, 2026. Registration is open: vapicon.ai

Please join us in welcoming Neil Zeghidour, Co-founder CEO of @GradiumAI, as a Speaker at VapiCon 2026! Neil founded Gradium after a decade of building and leading frontier generative audio teams at Meta and Google DeepMind. Being frustrated by slow and brittle voice assistants, he built the engineering teams that developed the first neural audio codecs and introduced the first audio LLMs, such as AudioLM, at Google. He later created Kyutai to launch Moshi, the world's first real-time, full-duplex conversational AI, and Hibiki, the first simultaneous speech-to-speech translation system. Today, Gradium is focused on helping developers build natural, real-time voice agents by providing ultra-low latency streaming APIs that transition these breakthroughs from the research lab to production. Join us in San Francisco this November as Neil and more industry leaders come together to delve into the future of Voice AI. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

We’re excited to welcome @CekuraAI as a Startup Sponsor of VapiCon 2026! Cekura is the reliability layer for voice agents, helping teams test before launch, monitor real production calls, and continuously improve. Its platform runs synthetic conversations, surfaces failures, reproduces issues in simulation, and helps teams validate fixes before the next release. We’re thrilled to have the Cekura team joining the Voice AI community at VapiCon. Join us November 11–12 at Fort Mason in San Francisco as the voice AI community connects, builds, and launches what’s next. Learn more and purchase your VapiCon 2026 ticket at vapicon.ai

We’re thrilled to announce that @kwindla, Kwindla Hultman Kramer, CEO at @trydaily, is returning as a speaker for VapiCon 2026! Kwindla is a network engineer, co-founder of infrastructure company Daily, and part of the team behind Pipecat, the open-source real-time AI framework. Join us in San Francisco this November as Kwindla and more industry leaders delve into the future of Voice AI. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

We're sponsoring Knowledgefest, Out-Of-Pocket's healthcare ops conference, October 16-17 in San Francisco! We'll be demoing and sticking around for the day. If you'll be there, come find us to talk about improving patient experience with voice AI!

You changed one line in the prompt. You have no idea what else was affected. A voice agent can answer the same caller three different ways. One good test call isn't evidence you can depend on for production. Swipe to learn best practices for testing voice agents so you can be confident every update works. Full write-up → docs.vapi.ai/test/voice-test…

VapiCon is 8 weeks away. Time to make the trip official! We'll be at Fort Mason's Festival Pavilion in San Francisco this November with the builders, product leaders, operators, and teams shaping the future of voice AI. Grab your ticket, book your travel, and plan your stay nearby before discounted hotel rates close on October 11. Recommended travel dates: November 10-13. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/ ✈️ Book your travel: vapicon.ai/book-your-stay Our Amazing Sponsors: Deepgram, Krisp, Oxylabs, AssemblyAI, pyannoteAI, Speechmatics, Cekura, Modulate, Phonic, Hamming AI, Roark, ai-coustics

We’re thrilled to announce that @seldo, Laurie Voss, Head of Developer Relations at @ArizeAI, is joining us as a speaker for VapiCon 2026! Laurie Voss is Head of Developer Relations at Arize AI, the leading company for AI observability and evaluations. He has been a developer for over 30 years and was co-founder of npm, Inc. He believes passionately in making the web bigger, better, and more accessible for everyone. Join us in San Francisco this November as Laurie and other leaders in the Voice AI discuss the future of the industry. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

Supporting a language and working in it are not the same thing. Catch us at HumanX Europe in Amsterdam at booth F19, to get into what that actually takes. Latency budgets. Interruption handling. What breaks when you go from one language to four. Plus: Vapi is hosting dinner at Shakerato, alongside Reson8 🇺🇸 🇬🇧 🇿🇦 who build speech recognition tuned to more than 20 European languages. 📍 RAI Amsterdam, Booth F19 📅 September 22-24, 2026 🎟️ Dinner RSVP, by approval: luma.com/d7jixwna?utm_source…

10,000 builders in Barcelona, and the best conversation is on a rooftop. Vapi is at AI Summit Barcelona. Come say hi at booth F5 with Deepgram and tell us what voice agents you're building in production. Then, join us at AZUL Rooftop in Barcelona for drinks and tapas at golden hour 🌇 📍 WTC Barcelona 📅 September 22-23, 2026 🎟️ RSVP: luma.com/aisummit-w7qr?utm_s…

We’re excited to welcome @Speechmatics as a Startup+ Sponsor of VapiCon 2026! Speechmatics builds the speech layer voice agents run on: real-time transcription that holds up in noisy, overlapping, heavily accented, real-world audio. It delivers high accuracy, sub-second latency, and speaker diarization that tracks who said what, even in messy multi-speaker conversations. With support for 55+ languages and dialects, Speechmatics powers everything from contact center automation to medical scribes and live media captioning. As voice AI moves from demo to production, Speechmatics is the reliable speech layer teams build on. Join us November 11–12 at Fort Mason in San Francisco as the voice AI community connects, builds, and launches what’s next. Learn more and purchase your VapiCon 2026 ticket at vapicon.ai #VapiCon #VoiceAI #AI

We’re happy to announce that Guillermo Rauch @rauchg, CEO of @Vercel, will be speaking at VapiCon 2026! Guillermo is the founder and CEO of Vercel, where he leads the company’s mission to enable the world to ship the best products. Originally from Argentina, Guillermo became a self-taught developer at the age of ten, and has been a passionate contributor to the open-source community ever since. He is the designer and co-author of Next.js, the world's most popular React framework, and is also the creator of socket.io and Mongoose.js. Before founding Vercel, he co-founded LearnBoost and Cloudup, where he served as CTO through its acquisition by Automattic in 2013. Come join us in San Francisco this November as Guillermo and other industry leaders dive into what’s next for AI and voice. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

We’re excited to welcome @pyannoteai as a Bronze Sponsor of VapiCon 2026! pyannoteAI gives machines the ability to understand every human conversation. It is the speaker intelligence layer for conversations—a framework that turns raw, overlapping, real-world audio into clean, labeled speaker metadata. Built on more than a decade of dedicated research, pyannoteAI’s diarization models identify who spoke, when, and how. Available as open source, through an API, or on private infrastructure, pyannoteAI pairs with any STT, TTS, or Voice AI model to make everything downstream work better. Join us November 11–12 at Fort Mason in San Francisco as the voice AI community connects, builds, and launches what’s next. Learn more and purchase your VapiCon 2026 ticket at vapicon.ai #VapiCon #VoiceAI #AI

🎙️ We’re thrilled to announce that Brooke Hopkins, Founder & CEO of @Covaldev, is joining us as a speaker at VapiCon 2026! Brooke began her career building voice technology at Google Assistant before leading evaluation infrastructure at Waymo. It was there that she saw the parallel between self-driving systems and conversational AI: both require rigorous testing and evaluation to perform reliably in the real world. Today, through Coval, Brooke is helping teams replace slow, manual testing with automated simulation and observability for conversational AI agents, expanding test coverage and giving teams more confidence to deploy. Join us in San Francisco this November to hear from Brooke and other industry leaders exploring the future of AI and voice. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket: vapicon.ai/

We’re excited to welcome @AssemblyAI back to VapiCon as a Silver+ Sponsor! AssemblyAI is the Voice AI infrastructure layer that millions of developers build on. Its speech-to-text, voice-agent, and speech-understanding models serve as critical infrastructure for companies including Granola, HeyGen, Ashby, and ClickUp. Through a single API, developers and product teams can build voice-driven applications ranging from AI meeting assistants and call analytics to real-time voice agents. We’re thrilled to have AssemblyAI joining us at VapiCon 2026! Join us November 11–12 as the voice AI community connects, builds and launches what’s next. Ready to join us? Learn more and purchase your VapiCon 2026 ticket at vapicon.ai. #VapiCon2026 #VoiceAI #SpeechTechnology #AssemblyAI

Yesterday we announced a rebuild of Campaigns, our batch outbound calling feature. Today, lead engineer Adham El Fawal discusses how we built it. How do you trust a system that dials people at scale on your behalf? Build the orchestration as its own engine, keep it out of the call pipeline, and test it with real calls.

We’re thrilled to announce that Klemen Simonic, CEO of @soniox_ai, will be speaking at VapiCon 2026! 🎙️ Klemen has been building intelligent systems at the frontier of speech, language, and large-scale AI since 2008. His work spans Stanford University, Google, and Meta, bridging cutting-edge research with technology deployed at global scale. Meet us in San Francisco this November to hear from Klemen and other industry leaders exploring the future of AI and voice. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket at vapicon.ai/

We’re proud to welcome @oxylabs_io as a Platinum Sponsor of VapiCon 2026! The next generation of AI and voice agents will need more than conversation; they’ll need access to real-world information. Oxylabs provides the infrastructure that enables AI agents to access and interact with the public web, retrieve fresh information, research, compare, and act on data in real time. Trusted by Fortune 500 companies, Oxylabs combines 175M+ ethically sourced IPs with advanced data-acquisition solutions to make the public web accessible at scale. Join us in welcoming Oxylabs to VapiCon 2026! Join us November 11–12 as the voice AI community connects, builds, and launches what’s next. Ready to join us? Learn more and purchase your VapiCon 2026 ticket at vapicon.ai. #VapiCon #VoiceAI #AI

🚨 Last Call for Early Bird Tickets to VapiCon 2026! 🚨 Early pricing 𝗘𝗡𝗗𝗦 𝗧𝗢𝗗𝗔𝗬, and tomorrow, tickets increase by $500. If you've been planning to join the frontier voice AI summit for builders, technologists, product leaders, and operators shaping what's next in voice AI, now is the time to lock in the lowest rate. Register now for VapiCon. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Get your ticket at vapicon.ai/ Sponsors: Deepgram, Krisp, Oxylabs, AssemblyAI, pyannoteAI, Speechmatics, Cekura, Hamming AI, Modulate, Phonic, Roark

We’re excited to announce that Davit Baghdasaryan, Co-Founder & CEO of @krispHQ, will be joining us for a Fireside Chat at VapiCon 2026! 🎙️ Davit has helped bring real-time Voice AI to more than 200 million devices and supported global brands as they work through communication barriers, friction, global deployments, and the push to scale human and agentic support and voice communications worldwide. His perspective comes from direct work with enterprises integrating AI into live workflows, with a focus on AI’s ability to reshape communication, augment human performance, and improve productivity without adding complexity. Join us in San Francisco this November to hear from Davit and other leaders shaping what’s next in AI and voice. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Early bird tickets are limited — get yours at vapicon.ai/

Save $500 before prices increase on September 1. Join us for two days of demos, workshops, technical sessions, and conversations on what's working and what's next in voice AI. 📍 Fort Mason, San Francisco 🗓️ November 11-12, 2026 🎟️ Save your seat: vapicon.ai/

How do you build reliable software in the age of LLMs? Scott Lowe, lead product engineer on Simulations, sat down to talk through it: how Simulations adds to Vapi's testing suite, checking both the decisions your agent makes and the outcomes it drives, what shipping the GA took, and why voice agents are testable with the same techniques that have always made software reliable.

We’re excited to announce that 𝗛𝗮𝗿𝗿𝗶𝘀𝗼𝗻 𝗖𝗵𝗮𝘀𝗲 @hwchase17, Co-Founder & CEO of @LangChain, will be speaking at VapiCon 2026! 🎙️ Harrison co-founded LangChain with the goal of making it easier to use LLMs to develop context-aware reasoning applications. LangChain also builds LangSmith, a platform for agent engineering. Prior to starting LangChain, Harrison led the ML team at Robust Intelligence and the entity linking team at Kensho. He also studied statistics and computer science at Harvard. Join us in San Francisco this November to hear from Harrison and other leaders shaping what’s next in AI and voice. 📍 Fort Mason, San Francisco 📅 November 11–12, 2026 🎟️ Early bird tickets are limited — get yours at vapicon.ai/

We’re excited to welcome @KrispHQ as a Platinum Sponsor of VapiCon 2026! Krisp Voice AI is real-time voice infrastructure that fixes conversations at the source. Operating at the audio layer, Krisp isolates voices, improves turn-taking, reduces accent friction, and breaks language barriers in live conversations before they reach any human agent, AI model, or downstream system. Join us November 11–12 as the voice AI community connects, builds, and launches what’s next. Ready to join us? Learn more and purchase your VapiCon 2026 ticket at vapicon.ai. #VapiCon #VoiceAI #AI

VapiCon 2026 is where the frontier of voice AI takes shape. Save your seat in the room with builders, technologists, product leaders, operators, and teams shaping what's next. Early Bird tickets are available now through August 31. After that, prices increase by $500 on September 1. Register now: vapicon.ai/ 📍 Fort Mason, San Francisco 🗓️ November 11-12, 2026 Registration Sponsored by @KrispHQ

We’re proud to welcome @DeepgramAI as the Diamond Sponsor of VapiCon 2026! Deepgram is helping shape the future of Voice AI, and we’re excited to have their team joining us in San Francisco for two days of conversations, connections, and innovation. Deepgram’s voice AI platform provides APIs for speech-to-text, text-to-speech, and full speech-to-speech voice agents. Over 200,000+ developers use Deepgram to build voice AI products and features. Build With Voice AI at Scale, from Transcription to Voice Agents. Thank you, Deepgram, for supporting the Voice AI community. We’ll see you at VapiCon! Get your early bird tickets before they run out at vapicon.ai

Your voice agent works in the demo. The question is what happens when a real caller introduces complexities that you didn’t test for. 𝗩𝗮𝗽𝗶 𝗦𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻𝘀 𝗶𝘀 𝗹𝗶𝘃𝗲. Build, test, iterate, and monitor in one platform: Native to Vapi, with full context of your agents: → AI testers with personalities + scenarios modeled on your real callers → tool mocks for failure paths you can't produce on demand → explainable pass/fail results with full transcripts Know your agent works before your customer finds out it doesn't. Read the blog: vapi.ai/blog/vapi-simulation…

Voice AI is becoming more human, and the builders behind it are heading to San Francisco this November. Join the sharpest engineers and builders for two days of demos, workshops, and conversations on where voice AI is headed next. Register now vapicon.ai/

Next Wednesday join us in SF at Dev Tools Demo night hosted by @scalekitinc You'll hear from Dhwani Trivedi and learn more about getting your voice agents to production. RSVP here: luma.com/0owahtma

The top of the Humanness Index™ changed twice in one week. A month ago, the leading score was 95. The gap to human is closing, and the names at the top keep changing. On a model-agnostic platform, the most human voice is a config change away. Read the blog: vapi.ai/blog/humanness-index…

Don't dump raw transcript output to your CRM. Vapi's Structured Outputs extract and validate against your schema post-call. You configure the schema and what to write to your CRM. Three keywords to configure: 𝗲𝗻𝘂𝗺: locks a value to a fixed set of options 𝗱𝗲𝘀𝗰𝗿𝗶𝗽𝘁𝗶𝗼𝗻: tells the model what "high" actually means. 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝗱: a miss returns null vs an error, your integration still has to check.

In one week join us for a livestream featuring @EdCharbeneau from @DeepgramAI as we dive into how to choose between models and leverage them on Vapi. We'll cover Nova-3, Flux, and Aura-2 and more. Bring your questions! RSVP: luma.com/vapi-44bz Thursday August 6 10:00am PST

"Build vs. buy" is too vague for voice AI. With Vapi you choose which layers to own: SIP trunk, STT, TTS, LLM endpoint, tools/webhooks, and defined post-call schemas. Bring the parts you need without building and maintaining the entire call stack.

Every voice agent = transcriber + LLM + voice. The hard part is picking the right ones. So we launched Vapi Model Intelligence, a bundle of two features: → Model Presets: Balanced, High Intelligence, Ultra Fast, Cost Saver. Pick one, ship. → Model Performance Metrics: data on every model, measured on real Vapi calls, not labs. Cost, latency, WER for transcribers, intelligence for LLMs, humanness for voices.

The best voice models now sound nearly human. One month of the Humanness Index, 11,000+ blind votes: → Grok TTS: 95 → real human: 100 The Humanness Index uses the same cloned voice, the same script, and a human recording in the set. Listeners are picking the top models almost as often as the real person. Full results: vapi.ai/blog/humanness-index… vote: humannessindex.vapi.ai

Barge-in is the opposite side of endpointing. Endpointing asks: 𝗶𝘀 𝘁𝗵𝗲 𝘂𝘀𝗲𝗿 𝗱𝗼𝗻𝗲 𝘀𝗽𝗲𝗮𝗸𝗶𝗻𝗴? Barge-in asks: 𝘀𝗵𝗼𝘂𝗹𝗱 𝘁𝗵𝗲 𝗮𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝘁 𝘀𝘁𝗼𝗽 𝘀𝗽𝗲𝗮𝗸𝗶𝗻𝗴? “yeah” should not cancel playback. “no, wait” probably should. background noise should not stop the assistant. For voice engineers, user audio is only the first signal. The system requires a policy for when to cancel assistant audio, when to ignore input, and how to recover after yielding the floor. In Vapi, the main tuning surface is stopSpeakingPlan: numWords → words required before stopping voiceSeconds → speech duration required before stopping backoffSeconds → recovery time before speaking again

Two days into blind voting of voice models on our Humanness Index™, and xAI's Grok TTS model is at the top of the pack. Its humanness score? 96, just 4 points under a real human voice (100). The Humanness Index takes one voice and one quote, clones it across every major model, then plays the results blind for real listeners to score. Hear it, cast your votes, and contribute to the leaderboard 👇 humannessindex.vapi.ai/

Layered in the startSpeakingPlan, by precedence 1. customEndpointingRules - overrides everything 2. smartEndpointingPlan - the prediction model 3. transcriptionEndpointingPlan - punctuation, timeouts

End of turn detection is a prediction. The model scores whether the caller is still speaking and maps that to a wait.

Your voice agent is talking over a caller who paused to pull up their account number. Raising the silence threshold could help, but it will also make the responses feel laggy. What you need to know is when the turn ended

You can feel whether a voice is a person or a machine before you can explain why. Today, we're launching the Humanness Index™, a live ranking of voice AI models, judged by you. Same voices, same quotes, different models. You pick the most human ones and even benchmark them against an actual human voice. 👉: humannessindex.vapi.ai/

"What's the latency?" is the wrong question to ask when your voice agent feels slow. Your users don't care about a benchmark when they feel: - dead air waiting for a response - the agent going quiet mid-reply - being talked over or not being able to interrupt - responses that seem to lag on more complex questions Each points to a different stage 👇

Have a listen! 🎧 piped.video/IWsx62OpAEE?si=ARKE…