Fish Audio, legally Hanabi AI Inc., has raised $52 million in seed funding led by Coreline Ventures and Capital Today, betting that voice will become the primary way people interact with AI models.
The company grew from an unlikely start. Co-founder Shijia Liao, a former Nvidia video researcher, built the first version in his bedroom on a single GPU, frustrated by the robotic synthetic voices of early systems. Today the platform claims more than 8 million users and over $21 million in annual recurring revenue, and its open Fish Speech project has gathered more than 31,000 GitHub stars.
Its technology spans text-to-speech and voice cloning across 83 languages, and can reportedly clone a voice from a five-second clip in under 15 seconds. In blind listening tests, the company says 67% of listeners preferred its flagship S2.1 Pro model over rivals.
Fish Audio serves everyone from indie developers to regulated healthcare and finance customers, offering on-premises deployments with HIPAA compliance and zero data retention. Next up are voice-native large language models, real-time speech-to-speech translation and deeper API integrations.
Source: SiliconANGLE.
Comments 0