Fixie.ai

Hey everyone, we want to share a fun holiday experiment that we've been building in the Fixie workshop the last couple of weeks: https://lnkd.in/eV62wwj5 https://hisanta.ai allows you to have real-time voice conversations with Santa and Friends. It's 100% powered by LLMs and built on top of Fixie.ai, including a new set of Voice APIs that we'll be releasing soon. We're really excited about what's possible with LLMs and Voice, and this experiment is just the first step for us. Oh, and if you're feeling a little more naughty-than-nice this year, check out https://hisanta.ai?nice=0 for a different cast of characters. We'd love for you to give it a try and let us know what you think!

Today we’re thrilled to introduce Ultravox v0.4.1, a family of open speech models trained specifically for enabling real-time conversation with LLMs. Ultravox is trained to consume speech directly (just like humans do!), which means no need for a separate automatic speech recognition (ASR) stage. This is the first step in enabling truly natural, fluent conversations with AI. Ultravox's speech understanding is the best in open-source and is quickly approaching the quality of GPT-4o. We're also opening access to Ultravox Realtime, a set of APIs and SDKs for building and deploying AI Voice Agents. Latency is comparable to GPT-4o Realtime at roughly 1/3 the cost (only $.05/min). Try the experience for yourself: http://demo.ultravox.ai Read more on our blog: https://lnkd.in/eMamBZ6Y

Fixie was recently featured in the Cerebral Valley deep dive: https://lnkd.in/gd2qF7HD

Our co-founder and CTO Justin Uberti was recently featured on the AI Inside Podcast talking about all things AI, Voice, and what we're up to at Fixie with https://ultravox.ai and https://ai.town. Give it a listen! https://lnkd.in/gc9uARMm

Really excited to see the word spreading about hisanta.ai. We're hard at work now building out our v2 system - if you're an AI researcher and the idea of building a single AI model that can hear, see, and speak sounds interesting, check out our new Head of AI job posting! https://lnkd.in/gjcAf3vG