Fish Speech at a glance
Checked 15 Oct 2026Inference code and weights for Fish Audio Drama 3 (35B-A3B MoE) and S2-Pro (4B): multi-speaker TTS, voice cloning from 10-30 s of audio, inline emotion tags. Not open source: code and Drama 3 use the Fish Audio Community License (commercial use only under USD 500k annual revenue, no hosted TTS services); S2-Pro weights are research and non-commercial only. Heavy hardware: Drama 3 needs at least 48 GB VRAM on NVIDIA; S2-Pro 24 GB or CPU. Licence updated 13 Oct 2026.
RepoQuarry. “Fish Speech.” Checked 15 Oct 2026. https://repoquarry.dev/projects/fish-speechHow much of ElevenLabs does it replace?
2 of 4 jobs. Covers multilingual TTS and voice cloning locally, but commercial use is capped by revenue and hosted TTS/cloning services need a separate licence.
For speech recognition, voice agents, see other ElevenLabs alternatives. You may also need to build an interface, an API server and hosting yourself.
How much of Descript does it replace?
2 of 3 jobs. Voice cloning and text-to-speech only; no transcript-based audio or video editing.
For transcript-based editing, see other Descript alternatives. You may also need to build an interface, an API server and hosting yourself.
Facts, with sources
Every answer links to where we found it. We read the documentation; we did not install it.
We provide Docker images for the API server on both GPU (CUDA126 by default) and CPU.speech.fish.audio · install ↗ 1d ago
This Agreement is intended to allow, free of charge, research and non-commercial uses of the Fish Audio Materials and, as set out in Section III, limited uses for a Commercial Purpose by Licensees whose Annual Revenue is less than the Revenue Threshold.github.com · LICENSE ↗ 1d ago
This model is licensed under the Fish Audio Research License. Research and non-commercial use is permitted free of charge. Commercial use requires a separate license from Fish Audiohuggingface.co · s2-pro ↗ 1d ago
regardless of Your purpose or Annual Revenue, no license is granted under this Agreement to offer a Hosted Service, which requires a separate written license agreement from Fish Audio.github.com · LICENSE ↗ 1d ago
Any other use of the Fish Audio Materials for a Commercial Purpose, and any Hosted Service, requires a separate written license agreement from Fish Audio.github.com · LICENSE ↗ 1d ago
GPU memory: at least 48 GB for Drama 3 (FP8), 24 GB recommended for S2-Prospeech.fish.audio · install ↗ 1d ago
System: Linux, WSLspeech.fish.audio · install ↗ 1d ago
pip install --extra-index-url https://download.pytorch.org/whl/cu129 -e ".[cu129]"speech.fish.audio · install ↗ 1d ago