TOOLORA / Local reading edition
Fish Speech
A deployable speech project using Drama 3 for long multi-character audio, with emotion instructions, iterative revision and local edits; also runs S2-Pro.
Overview
Fish Speech provides command-line, API-server and WebUI deployment code for speech models, distinct from the hosted Fish Audio platform. Current Drama 3 targets multi-character scenes and long content, with natural-language emotion tags, iterative revision and local speech editing; the same runtime also supports previous-generation S2-Pro. Drama 3 ships with a Community License that conditions commercial use and hosted services. S2-Pro weights retain a Research License, so downloadable code does not imply unrestricted commercial model use.
Official sources and references
- Official websitehttps://github.com/fishaudio/fish-speech
- Documentationhttps://huggingface.co/fishaudio/drama-3
- Documentationhttps://speech.fish.audio/
Descriptions summarize public information; consult official websites for features, pricing and regional availability.