Fish Audio
Expressive Text-to-Speech and Voice Cloning
Start new thread

Fish Audio Drama 3 - AI voice acting you can direct, not just generate

by
•
Drama 3 is Fish Audio's latest AI voice model, built for performances you can direct. Turn scripts into expressive speech with natural-language control over emotion, pacing, whispers, laughter, and pauses — even within a single sentence. Create multi-character scenes, audiobooks, game dialogue, and narration across 83 languages. Available on the web, mobile, and API, with open weights for self-hosting under Fish Audio's community license.

Add a comment

Replies

Best
Hunter
📌
Fish Audio just released Drama 3, and the voice samples are worth a listen. Instead of just choosing a voice and generating speech, you can actually direct how a line is performed — where someone whispers, laughs, pauses, or changes emotion halfway through a sentence. It also handles multi-character dialogue, supports 83 languages, and has downloadable model weights. They're working on a feature that lets you change a single word without regenerating the whole recording, which I'm particularly interested in. Pretty cool for anyone making audiobooks, games, or character-driven videos.

How natural does the voicesound compared to real human speech

I saw the uncle roger ad before seeing the launch here. I don't know what is SOTA but voice cloning took like half a minute! That's incredible (and a little scary!)

Can I use the voices in a paid client video on the free plan, or only on a paid one?

Controlling a pause or an emotion change inside one sentence feels much closer to directing a performance. That's the level where a line can be technically correct and still sound wrong.