Podcastle Introduces Advanced Text-to-Speech Model Featuring Over 450 AI Voices

Podcastle, a podcast recording and editing platform, has launched its new AI-driven text-to-speech model, Asyncflow v1.0, offering over 450 AI voices for text narration, alongside an API for developers.

Podcastle, already a big name in podcast recording and editing, is now throwing its hat into the AI text-to-speech ring with Asyncflow v1.0. And let me tell you, it’s not playing around—boasting over 450 AI voices that can turn your boring old text into something that actually sounds alive. For the tech-savvy folks out there, there’s an API to weave this magic into your own apps, because who doesn’t love more voice options?

What’s cool about Asyncflow v1.0? It was built to keep those pesky training and inference costs low, giving Podcastle a leg up in this crowded market. They’re rubbing shoulders with the likes of ElevenLabs and Speechify, all racing to turn text into speech that doesn’t sound like your GPS. From ads to e-learning, the possibilities are endless.

Arto Yeritsyan, Podcastle’s brainchild, spilled the beans to TechCrunch about the rollercoaster of creating a text-to-speech model that’s both wallet-friendly and top-notch. Thanks to a hefty $13.5 million Series A funding, they’ve managed to pull off this balancing act—less data needed, but no skimping on quality.

And here’s the kicker: Podcastle is turbocharging its voice cloning. With just a snippet of your voice (yes, really, just a few seconds), thanks to their Magic Dust AI, you can have a digital twin. Sure, it might sound a bit like a robot at first, but they’re on it. They’re even nudging users to play around with different samples to see what sticks.

Bottom line? Podcastle’s packing a serious toolkit—audio, video, podcasts, and now AI narration. It’s shaping up to be a one-stop-shop for digital content creators looking to stand out in a sea of sameness.