Fish Audio
AI voices and voice cloning from a short audio sample.
- Made by
- Fish Audio
- Category
- Voice
- Last change
- Oct 13, 2026
Changelog
October 2026
- New model
Fish Audio released Drama 3, an open-weight voice model that controls the tone of single words and shifts voice mid-sentence, on web and API.
September 2026
- New feature
Fish Audio's speech-to-text model now identifies different speakers, understands their feelings and labels emotion cues like [laughter] in transcripts across 83 languages.
- New model
Fish Audio launched a preview of Drama 3, a voice model steered by plain-language tone, pacing and character descriptions, available through its API.
- New feature
Fish Audio MCP added image and video generation with cost shown before creating, for users of agents like Claude, Codex, Cursor and Windsurf.
- Availability
Fish Audio S2.1 Pro is now available in Europe and Australia through SLNG, with in-region GPU infrastructure and 83 languages.
August 2026
- Availability
Fish Audio released its voice studio app for Android, following the iOS launch.
- Launch
Fish Audio launched an iOS app with voice discovery, multi-speaker dialogue in 80+ languages, emotion tags, and voice cloning or design from a prompt.
