Voice and music · by Fish Audio

Fish Audio

Clone any voice from a few seconds of audio.

text → speech · speech → profile · text → profile
Way in
Official APISign up yourself, api key
To start
Free tiers2.1-pro-free model at $0; app Free plan 8k credits
Runs
In the cloudAlso through OpenRouter
Checked
Nov 3, 2026Prices as of that day

What it can do 5

Synthesize expressive speech (Drama 3 / S2.1 Pro)concurrency 5-50 by spendtext → speech$15/1M UTF-8 bytes (S2.1 Pro); $50/1M Drama 3 ($10 promo until 2026-11-12); s2.1-pro-free $0
Clone a voice from a short samplespeech → profilenot priced separately
Design a voice from a descriptiontext → profile$0.01/request
Transcribe audiospeech → text$0.36/hour
Self-host the Drama 3 35B MoE TTS weightsabout 48 GB+ NVIDIA VRAM in FP8 (not runnable on this 16 GB Mac)text, speech → speechfree under Fish Audio Community License for orgs under $500k revenue; licence needed above that or for hosted services

Does the same job

Prices and limits as of Nov 3, 2026, from Fish Audio’s own pages. They change; check before you rely on them.Not affiliated with Fish Audio. Not advice.Something wrong?
Made by Jasper Burns

A short note when I ship something new, about once a month. One list for all my projects.