GitLab Transcend - Intelligent orchestration
Mini toolsDramabox
By Resemble AI
DramaBox is a prompt-driven expressive TTS model from Resemble AI, released on Hugging Face under the LTX-2 Community License. It is built on Lightricks’ LTX-2.3 audio branch and fine-tuned as an IC-LoRA on a 3.3B audio-only diffusion transformer with flow matching, conditioned on Gemma 3 12B text embeddings. Its prompt format separates dialogue inside quotes from stage directions outside quotes, letting the model control speaker identity, emotion, delivery, laughs, sighs, breaths, pauses, and transitions. It also supports optional voice cloning from a 10-second or longer reference clip and applies an imperceptible Resemble Perth watermark by default.
New
Multimodal
Gen 3
Overview
DramaBox is Resemble AI’s expressive text-to-speech and voice-cloning model for turning screenplay-style prompts into emotional spoken performances.
About Resemble AI
Resemble AI is an Artificial Intelligence company that offers a voice generator with voice cloning for text to speech.
View Company ProfileTools using Dramabox
No tools found for this model yet.
Last updated: May 22, 2026
