Fish Audio

Fish Audio is a text-to-speech and voice cloning platform built around an open model line and a community voice library. You can clone a voice from a short sample, browse thousands of voices other users have published, or use the models directly.

What it offers

• Voice cloning from a brief recording, capturing timbre and delivery rather than just pitch
• A public voice library with thousands of community-contributed voices across many languages and styles
• Multilingual synthesis with natural prosody, including cross-lingual use of a cloned voice
• Open models. The underlying speech models are published openly, so they can be self-hosted or fine-tuned
• An API with streaming support for real-time applications

Where it fits

Narration for video and audiobooks, character voices for games, accessibility tooling, and voice interfaces where per-character costs matter. The open-model approach keeps pricing well below the premium end of the market while staying competitive on quality.

Free credits are available for evaluation; paid usage is metered per character with subscription tiers for volume. Cloning someone else's voice requires their consent, and the platform's terms reflect that.