HeyGen Voice HeyGen Voice is HeyGen’s first in-house text-to-speech model, built for authentic cloned-voice performance that stays natural and expressive rather than flattened. It is live in the HeyGen product and via API, and independently ranked #1 on the Artificial Analysis Controlled Voice Arena in blind tests against major TTS models.
Visit productFeatured media
Loading media…
View on XCosigns Received
“really pumped to see our own voice model finally out, we were testing it internally/externally for a long time and it beats the previous best voice models by a lot, I remember choosing our own voice model in blind side by side for Spanish voices, it's that crazy and cheap 🤣”
“Heygen Voice 1 is the best version of my voice I've ever heard. It's not just how it sounds but also the intonations and tone it uses by default Congrats to the team and here's a little video in @HyperFrames_ Studio to showcase it”
“Nothing is more personal than your voice. Every AI voice model we tried either flattened it or made you sound like someone else. So we built our own, for authentic performance that sounds like you at your best. HeyGen Voice is live, ranked #1 on @ArtificialAnlys!”
“i have a year’s supply of @LiquidDeath , a new fridge, & i needed an excuse to try @HeyGen voice this cost less than $0.1 to generate the narration created my own custom voice with a one sentence prompt, then fed it a script”