Fish Audio S2 - Real Expressive AI Voices
by•
We've open-sourced Fish Audio S2, a new generation of expressive TTS that lets you direct voices with natural language. Add cues like [whisper] or [laughing nervously], generate multi-speaker dialogue in one pass, and create scary-real voices across 80+ languages.
Replies
Sway
Pricing itself makes a huge difference compared to competitors. And the quality is on par with most of "high end" TTS models
Very cool! In voice AI, the lack of emotions is the main problem.
I've tried Fish Audio and the voice cloning quality genuinely impressed me — it sounds almost identical to the original. Big thanks to the team for building this! Can't wait to see what S2 brings.
I love the simplicity, its so easy to use and gives great result that removes half of the stress of your audio projects
Fish Audio es una joya tecnológica por su fidelidad y velocidad, permitiendo una clonación impecable. Sin embargo, quitar la temperatura es un error garrafal: le quita el alma al relato. Al automatizar la emoción, transforman una herramienta expresiva en un robot plano, privándonos del caos necesario para transmitir sentimientos reales.
This is a top tier quality, it's my new favorite 🤯. And it's only a 4.5B model too 😯 incredible!
Great blog post too. I can't wait to get some time to go over the technical report you released.
I've always wanted a TTS that can do a bunch of tags while preserving fairly good similarity and flow. Fish audio s2 pro surprised me with all these and... I'm loving it! Hope to explore more usage cases with s2-pro model!
Been using this to enhance my DND games.. its transformed it! allocating voices to actual NPC parts is awesome.. highly recommended and much fairer priced then Elevenlabs - highly recommended!!!
Congrats on the launch! I'm curious: if I’m building a real-time voice agent where latency and fine-grained emotion are dealbreakers, what specific benchmarks or features make Fish Audio a better bet than ElevenLabs right now?