1-844-HEY-VAPI – voice AI platform for developers
Hey HN, I’m Jordan, founder of Vapi. For the past 1.5 years, we've been building voice infrastructure designed to help developers easily create, test, and deploy voice agents. We’re everything between raw models (like the realtime API) and end-users interacting with voice agents over telephony or within applications. We’ve been around for about 1.5yrs. Vapi was born out of a personal frustration—I wanted to build my own voice agent but found existing solutions extremely limited. Back then, there were only a handful of voice AI startups. Since March 2024, we've witnessed a Cambrian explosion…
In plain words
Vapi is a voice infrastructure platform that enables developers to build, test, and deploy voice agents. It sits between raw AI models and end-users, handling voice interactions across telephony and applications. The platform addresses the gap between basic model APIs and production-ready voice solutions, allowing developers to create voice agents for business applications like customer support, payment processing, and authorization workflows without building infrastructure from scratch.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey HN, I’m Jordan, founder of Vapi. For the past 1.5 years, we've been building voice infrastructure designed to help developers easily create, test, and deploy voice agents. We’re everything between raw models (like the realtime API) and end-users interacting with voice agents over telephony or within applications. We’ve been around for about 1.5yrs. Vapi was born out of a personal frustration—I wanted to build my own voice agent but found existing solutions extremely limited. Back then, there were only a handful of voice AI startups. Since March 2024, we've witnessed a Cambrian explosion of new entrants. In the early days, the biggest challenge was always latency. But as models have rapidly improved and funding has flooded into the space, voice agents are now trusted for business-critical applications like prior authorizations, payment processing, and customer support. Today’s challenge is reliability: can these voice agents consistently perform in production? What we’ve learned deploying 1000s of them: Yes, but it's really hard. Why? Real-time is unweildy—coordinating 10+ third party real-time models, handling diverse, unconstrained inputs and outputs, with no latency budget for reasoning and reflection. Voice interactions don’t follow a linear flow like text, making determinism a challenge. Getting it right takes a ton of tuning and iteration. Our approach does not simplify, it assumes it will be hard. We’re built for developers. We expose extensive configuration options through an API-native platform, including custom models, 100+ integrations, and detailed call testing and monitoring features. -- Tldr; it works! - We achieve performance equivalent to humans for use cases with clear user intent (appointment bookings, support, etc.). - We've handled over 44M voice calls across support automation, healthcare intake, and logistics. - We reliably manage long-running jobs (10–60+ minutes) with 99.95% uptime. Questions we’re still exploring: - When you have a complex product with 1000s of nobs, how do you help developers wield them, without heavy consulting? - How do you detect and eliminate bad actors when the line can be blurry / not enough context to make the call on what is fraud and what isn’t? - How do you guarantee reliability when we are built on services that are less reliable than industry standard (e.g. OpenAI posts < 99.9% uptime) -- You can try a live agent by calling: 1-844-HEY-VAPI. Yes, that’s my voice. Docs: https://docs.vapi.ai/introduction Let me know if you have any questions about voice; the market or the tech. Happy to poke the eng team to chat on architecture, etc. if it’s of interest. PS- we are also launching on product hunt today. we got an awesome team working really hard to make that happen, so pls support, thanks!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, April 2025
the whole month →- IB
Hi everyone, I built PyXL — a hardware processor that executes a custom assembly generated from Python programs, without using a traditional interpreter or virtual machine. It compiles Python -> CPython Bytecode -> Instruction set designed for direct hardware execution. I’m sharing an early benchmark: a GPIO test where PyXL achieves a 480ns round-trip toggle — compared to 14-25 micro seconds on a MicroPython Pyboard - even though PyXL runs at a lower clock (100MHz vs. 168MHz). The design is stack-based, fully pipelined, and preserves Python's dynamic typing without static type restrictions.…
Dev tools · 2025 · runpyxl.com
- UC
Life & fun · 2025 · filiph.github.io
- IB
https://the-pocket.github.io/Tutorial-Codebase-Knowledge/
AI · 2025 · github.com
