Fish Audio is the product arm of Hanabi AI Inc., founded by a young team led by CEO Shijia Liao, an active open-source AI developer. The company grew out of the widely used Fish Speech open-source TTS project and has become one of the most credible challengers to ElevenLabs, scaling annualized revenue from $400K to over $5M in early 2025 while growing to hundreds of thousands of monthly users.
Products
The platform offers emotionally controllable text-to-speech with inline emotion tags, voice cloning from just 15 seconds of audio, speech-to-text with multi-speaker detection, a voice changer, audio translation and a Story Studio. Its OpenAudio S1 model, launched in 2025, was positioned as "the world's first AI voice actor" with real-time emotional and tonal control. A community library hosts over 2 million voices.
Developers and funding
A low-latency REST API with pay-as-you-go pricing and SDKs serves more than 20,000 active developers, and the core models remain open source on GitHub. The company reportedly closed a $30M round at a $500M valuation to scale open-source AI for creative production.