What began as a co-founder’s personal passion has evolved into a venture-backed startup as the company passes its first anniversary, having quickly gained a user base of creators, developers, and enterprise customers.
Fish Audio is transforming an open-source passion project into one of the fastest-growing companies in the competitive voice artificial intelligence market.
The Palo Alto, California-based company announced that its $52 million seed funding round was led by Coreline Ventures and Capital Today. Additional participants included 359 Capital, Play Time, HF0, 645 Ventures, Parable, Carya Venture Partners, Alphalist Partners and several angel investors.
The investment comes as Fish Audio celebrates its first anniversary, having reported significant early growth. Since its founding in 2025, the company quickly reached $21 million in annual recurring revenue. Its 8 million-plus users include creators, software developers, and enterprise customers.
From Open-Source Project to Voice AI Company
Fish Audio originated as a personal project developed by Co-Founder and Chief Scientist Shijia Liao, a former NVIDIA video researcher and longtime anime and VTuber enthusiast.
Frustrated by synthetic voices that sounded monotonous or robotic, Liao began training voice models from his bedroom using a single gaming graphics processing unit. His work eventually became Fish Speech, an open-source voice AI project. It quickly gained attention among independent developers, video game designers and digital content creators.
Fish Speech has since earned more than 31,000 stars on GitHub, making it one of the platform’s most popular open-source voice projects. Early developer interest helped form the foundation for Fish Audio’s commercial platform.
Today, Fish Audio provides real-time text-to-speech technology, voice cloning and AI-powered voice agents. Its technology can clone a voice from a five-second recording in approximately 15 seconds and supports more than 83 languages.
The platform also offers word-level emotional control through more than 15,000 natural-language instructions. This gives developers and creators greater control over how generated voices communicate tone, intent, and personality.
“We built Fish Audio because we wanted voice AI that sounded human, not chunky or robotic, and we wanted that quality to be accessible at any scale,” said Rissa Cao, co-founder and CEO of Fish Audio.
Fish Audio Targets Enterprise Voice AI Market
Fish Audio is also positioning its technology for enterprise adoption, including organizations operating in highly regulated industries.
The company offers on-premises deployment, zero-data-retention policies and configurations designed to meet HIPAA compliance requirements. These capabilities allow companies to adopt voice AI while maintaining greater control over sensitive information, infrastructure and security requirements.
Fish Audio said its latest model, S2.1 Pro, was preferred by nearly 67% of listeners over leading competitors during blind listening tests.
Osuke Honda, managing partner at Coreline Ventures, said voice is becoming a primary interface for artificial intelligence applications. He added that Fish Audio’s performance, multilingual capabilities, emotional expression and pricing have helped it gain traction among creators, developers and enterprise customers.
The company’s customer base includes HeyGen, Retell, LiveKit, OpenArt, Telnyx and Sanas.
Funding Will Support Full Audio-Native AI Stack
Fish Audio plans to use the $52 million funding round to expand beyond text-to-speech technology and build what it describes as a complete audio-native AI stack.
Future development areas will include voice-native large language models, speech-to-speech technology and additional audio AI products. The company will also expand its enterprise sales organization and strengthen its developer tools, integrations and partnerships.
Fish Audio currently works with voice and communications technology providers including LiveKit and Retell, helping developers incorporate expressive AI voices into customer-service platforms, content tools, games and other interactive applications.
Through the end of August 2026, Fish Audio is making its S2.1 Pro model available to developers at no cost through its application programming interface.
With fresh capital, rapid revenue growth and millions of users, Fish Audio is seeking to establish itself as a leading platform as voice becomes an increasingly important way for people to interact with artificial intelligence.
