Fish Audio has launched Drama 3, which it calls the most emotive and controllable text-to-speech model in the world. The company also says it is open-weight and backed by a $52M seed round with 15 million-plus users.
Those are company claims, so here is what they mean in practice.
More on this topic: TechnoSports Fish Audio coverage.
Fish Audio Drama 3: Key Details
| Claim | Detail |
|---|---|
| Model | Drama 3 |
| Type | Text-to-speech, open-weight |
| Control | Tone control per word, mid-sentence voice shifts |
| Funding claim | $52M seed round |
| Users claim | 15M+ |
| Promo | 1 month free for commenters within 48 hours; first month free for enterprises |
What Drama 3 Claims to Do
Fish Audio says you can control how even a single word should sound, change its tone and shift your voice mid-sentence. That level of control suits audiobooks, games, ads and dubbing, where a flat voice breaks the mood.
The post, which had over 8 million views, also says the model is open-weight, meaning developers may be able to run it themselves. The video featured the creator Uncle Roger reacting to the demo.
The Promo Offer
The company asked viewers to comment DRAMA within 48 hours for one month free, and said enterprises get their first month free. These offers come from the post and may change. We have not linked the promotional short links, so use the official site.
Learn about the basics of speech synthesis or visit Fish Audio for official details.
Why Creators Should Care
Voice control down to a word lets creators fix a line without re-recording. Open weights also add choice, since small studios can host the model privately for sensitive scripts.
See our text-to-speech coverage for other launches.
Questions Before You Adopt It
Check the licence for commercial use, the supported languages and the hardware needed to run the weights. Also test how the model handles long scripts, since emotion control can vary between short demos and full chapters.
Voice cloning raises consent and misuse questions, so ask how the platform handles voice permissions before using real people’s voices.
What Is Not Confirmed
- The $52M seed round and 15M user figures, which come from the company post only.
- The licence for the open weights.
- Languages supported, including Indian languages.
- Independent quality comparisons.
Quick Answers
What is Drama 3?
An open-weight text-to-speech model from Fish Audio with fine emotion control.
Is it free?
The post offers a free first month in a limited promotion.
Can I control single words?
Yes, per the company claim.
Source: Fish Audio on X, Fish Audio official site, Speech synthesis on Wikipedia.
Use Cases Where Word-Level Control Helps
Audiobook publishers can fix a single line without re-recording a chapter. Game studios can give non-player characters a range of moods from one voice. Advertisers can localise a script while keeping the same tone.
The same power raises misuse risks such as fake voices. Look for consent tools, watermarking and clear terms before using any voice that sounds like a real person.
Bottom Line
Drama 3 sounds like a strong tool for creators if the claims hold. The funding and user numbers come from the company itself, so check independent reviews and the licence before relying on it.
Test it on your own scripts, and read the terms on voice rights before publishing anything made with it.
