Fish Audio models now available on Vercel AI Gateway for free
Fish Audio's four models—including low-latency voice cloning and 80-language TTS—are free on Vercel AI Gateway through September 18. Practitioners building audio features just got a no-cost ramp.
Why it matters
Vercel AI Gateway integrates Fish Audio's speech and transcription models with free access for 30 days, lowering the barrier for developers to add production-grade audio to applications without upfront cost.
The key facts
15 to knowFour Fish Audio models now on Vercel AI Gateway
Free access through September 18, 2026 (30-day offer)
Models include text-to-speech (s2.1-pro, s2-pro, s1) and speech-to-text (transcribe-1)
fish-audio/s2.1-pro: low-latency voice cloning from reference recording
fish-audio/s2-pro: supports 80+ languages with inline style tags
fish-audio/transcribe-1: word-level timestamped segments
Regular pricing: $15/million characters (TTS), $0.36/hour (STT)
Available via Vercel AI SDK v7 with generateSpeech() and transcribe() functions
Playground available for no-code testing in browser
Four Fish Audio models available: s2.1-pro (low-latency voice cloning TTS), s2-pro (80-language TTS with inline tags), s1 (emotional/tone-aware TTS), transcribe-1 (word-level timestamped transcription)
Free access through September 18, 2026 on Vercel AI Gateway
Regular pricing: $15/million characters for TTS, $0.36/hour for speech-to-text
Integrated into Vercel AI SDK 7 with generateSpeech() and transcribe() methods
Word-level segmentation available in transcription output
Models available via Vercel's playground (no-code access) and programmatic API
Go to the source
Vercel Blogvercel.com
Publisher excerpt: 's audio models are now available on AI Gateway. To celebrate the launch, every model is free on AI Gateway for the next 30 days, through September 18.Fish AudioFish Audio Four models from Fish Audio are available, including their latest text-to-speech model: How to use models during the offer…