Fish Audio's Drama 3 text-to-speech model now available on AI Gateway

Fish Audio's Drama 3 text-to-speech model is now available on AI Gateway. The model is currently in preview and allows users to direct voice performances using natural-language instructions embedded directly in scripts, making it suitable for voiceovers, character dialogue, and audio storytelling.

Directing Voice Performances

Users can control emotion and pacing by placing instructions in square brackets, inserting sounds and pauses, or wrapping words in tags like <whisper> and <emphasis> to alter delivery. These directions can be written in a different language than the spoken text. Drama 3 also supports multi-speaker dialogue, assigning a separate voice to each character.

Access and Pricing

To use the model, specify fish-audio/drama-3 as the model name. A promotional discount of 80% off is available for 30 days, ending on November 12. AI Gateway provides a unified API for calling models, tracking usage, and configuring retries and failover, with no markup on provider pricing or platform fees for inference.