Hanabi AI has introduced OpenAudio S1, a new vocal synthesis tool that they say treats emotion as the core of AI voice experience, allowing users to direct the voice performance, adjust tone, pacing, and feeling as naturally as working with a human actor.

Ahead of launch, OpenAudio S1 was submitted to Hugging Face’s TTS Arena, which is sort of like the old ‘Hot or Not’ site, but for text-to-speach (TTS) vocal synthesis. Instead of ranking photos of people, you vote on head-to-head comparisons of the results from two different vocal synthesis engines.

Here’s how TTS Arena works:

• Enter your text and select ‘synthesize’.
• Listen to two different TTS models synthesize the same content.
• Vote for the model that sounds better.
• Track overall model rankings on the leaderboard. You can also create an account on the site and create your own leaderboard.

Hanabi AI shared this with us because, of course, OpenAudio S1 is currently the leading text-to-speech engine on TTS Arena.

OpenAudio S1 lets you ‘tag’ your script with a variety of markers, including Emotion, Tone, and markers for things like laughing or sighing. The system uses these tags as hints for generating more realistic vocal synthesis results.

“The future of AI voice-driven storytelling isn’t just about generating speech—it’s about performance,” said Shijia Liao, founder and CEO of Hanabi AI. “With OpenAudio S1, we’re shaping what we see as the next creative frontier: AI voice acting.”

For more info on OpenAudio S1, see their launch blog post.

If you give TTS Arena or OpenAudio S1 a try, leave a comment, and let us know what you think of the current state of voice synthesis!

Pricing & Availability:

TTS Arena is free to try or to view their leaderboard. OpenAudio S1 is available now at Fish.Audio, priced at $15/month or $120/year. You can also check out the OpenAudio open source TTS repo on Github.