Paid s2.1-pro is $15 per million UTF-8 bytes
Fish Audio S2.1 Pro Text-to-Speech API Is Free With No Hard Character Cap
Free S2.1 Pro text-to-speech API with no hard character cap under fair use
Fish Audio has opened its flagship S2.1 Pro voice model to developers for free. Set the model to s2.1-pro-free in your API calls and there's no hard character cap under fair use. The free window runs through November 30.
- Eligibility
- Any developer with a Fish Audio account; products above $1M ARR should contact Fish Audio first
- Benefit duration
- Free through November 30, 2026
- Claim by
- 2026-11-30
The briefing
What you get
S2.1 Pro is Fish Audio's best text-to-speech model. One model covers 83 languages, supports voice cloning and multi-speaker dialogue, and lets you steer delivery with bracket tags such as [happy] or [whispering]. The free model, s2.1-pro-free, is the same model as the paid s2.1-pro, which normally costs $15 per million UTF-8 bytes.
Free access runs through November 30, 2026. There is no hard character cap under the fair-use policy, and you don't need to add a card or top up. The free tier comes without latency or uptime guarantees and is meant for development, testing, prototyping and smaller businesses. Requests may be used to improve the model, and products above $1M ARR should contact Fish Audio first.
How to use it
- Sign up at fish.audio and create a key on the API Keys page. The key is shown only once, so save it right away.
- Send a POST request to https://api.fish.audio/v1/tts with Authorization: Bearer YOUR_KEY and a model: s2.1-pro-free header. Put the text you want spoken and the output format (for example mp3) in the body.
- To pick a voice, add a reference_id from the public voice library, or upload a short recording to clone your own voice.
After November 30, switch the model to s2.1-pro to keep going on pay-as-you-go pricing.
See the S2.1 Pro free API announcement and API pricing for details.