Summary
On August 19, 2026, Vercel added Fish Audio's speech and transcription models to the Vercel AI Gateway, accessible through a single API key with built-in spend monitoring and offered at no cost for the first month.
What changed
Vercel AI Gateway added four Fish Audio models (the s2.1-pro, s2-pro, and s1 text-to-speech models plus the transcribe-1 transcription model) to its catalog. Developers can call them through the same single Gateway API key with integrated spend monitoring, and every Fish Audio model is free for 30 days through September 18 when called with a -free suffix.
Why it matters
It extends the AI Gateway beyond text and coding models into voice, positioning Vercel as a single metered endpoint for multimodal model access and using a free introductory month as an acquisition lever the way it has for other newly added providers.
Evidence excerpt
Speech and transcription models from Fish Audio became available at no cost for the first month, requiring only a single API key with integrated spend monitoring.