Fish Audio
Text-to-speech and voice cloning API with 83 languages and ~70ms latency
Important Notes
Same S2.1 Pro model as the paid tier — normally $15 per million UTF-8 bytes — free through August 31, 2026 with no hard character cap. Voice cloning and all 83 languages included.
Full details & quickstart →No SLA or latency guarantee, and requests may be used to improve the model. Built for prototyping and evaluation, not production.
Open to everyone — no student verification needed. Sign up, create an API key, then set the model header to s2.1-pro-free.
More in this category
Merge Gateway
$10/mo in free LLM credits, any major model
Brik
3 months Pro free
B.AI
Up to 1M free credits on signup, no card required
HyperWrite
50% off monthly subscription
More in AI & Machine Learning
Offers verified regularly. Student status must be maintained. Terms vary by provider and region. Not affiliated with listed companies.