Fish Audio has raised a $52 million seed round to accelerate development of its real‑time voice AI platform for creators, developers and enterprises. The Palo Alto startup—reporting roughly $21 million in annual recurring revenue and more than 8 million users after its first commercial year—says it will expand beyond text‑to‑speech into voice‑native language models and speech‑to‑speech systems.
Fish Audio raises $52M to push Voice AI
The round was led by Coreline Ventures and Capital Today, with participation from 359 Capital, Play Time, HF0, 645 Ventures, Parable, Carya Venture Partners, Alphalist Partners and several angel investors. Fish Audio says the capital will beef up enterprise sales, developer tools and integrations, including partnerships with LiveKit and Retell.
Fast, multilingual and expressive tech
The company highlights several technical benchmarks: voice cloning from a five‑second sample in about 15 seconds, support for more than 83 languages, and fine‑grained emotional control using over 15,000 natural‑language commands that can modulate expression word‑by‑word. Fish Audio reports that its S2.1 Pro model was preferred by about 67% of listeners in blind listening tests.
Open‑source roots and commercial traction
Fish Audio grew out of the open‑source Fish Speech project led by cofounder and chief scientist Shijia Liao, which has accumulated more than 31,000 stars on GitHub. That foundation helped attract a community of game developers, indie builders and digital creators and eventually spun out into a commercial product that reached the cited usage and revenue milestones.
Security, deployment and customers
Mindful of regulatory and privacy requirements, Fish Audio offers on‑premises deployment, no‑data‑retention configurations and HIPAA‑compatible plans for customers with strict compliance needs. The company lists customers including HeyGen, Telnyx, OpenArt, Sanas and other voice and media players.
Product ambitions
Beyond TTS, Fish Audio plans to invest in voice‑native language models, speech‑to‑speech systems and other advanced audio capabilities. The stated goal is to deliver realistic, multilingual, emotionally controllable synthetic voices for use cases ranging from individual creators to large enterprises.
“We wanted a Voice AI that sounds human and is accessible at any scale… Eight million users have already experienced it,” said Rissa Cao, cofounder and CEO. Osuke Honda, managing partner at Coreline Ventures, added that Fish Audio quickly differentiated itself on performance, multilingual support, emotional expression and cost.
The funding should help Fish Audio accelerate its transition from a TTS solution to a full audio platform while strengthening commercial traction and technical integrations.
