Live1 models

Fish Audio API (S2.1 Pro)

The Fish Audio API on Muapi gives you Fish Audio S2.1 Pro, a high-fidelity text-to-speech model for long-form narration. The Fish Audio S2.1 Pro API accepts up to 10,000 characters in one request and returns an MP3, billed at $15 per 1M UTF-8 bytes (about $0.015 per 1,000 English characters).People also search for this as the Fish Speech API. Call the fish-s2-1-pro endpoint with your Muapi key, then poll the request ID or use a webhook to collect the audio.

Get a Fish Audio API Key Try Fish Audio S2.1 Pro Text-to-Speech API examples and comparison on GitHub

The Fish Audio S2.1 Pro API is live on Muapi

Send up to 10,000 characters, poll the returned request ID, and download an MP3.

Fish Audio S2.1 Pro

Long-form text to speech, up to 10,000 characters per request1 models

TTSNewFeatured

S2.1 Pro

Fish Audio S2.1 Pro

High-fidelity long-form text-to-speech: up to 10,000 characters per request, voice_fish_narrator, MP3 output, billed at $15 per 1M UTF-8 bytes.

Text → audio, up to 10,000 chars

$15/1M UTF-8 bytes

Try Model

What is the Fish Audio API?

The Fish Audio API on Muapi (endpoint fish-s2-1-pro) runs Fish Audio S2.1 Pro, a text-to-speech model built for long-form narration. A single request can carry up to 10,000 characters in prompt, so whole scripts, chapters, and articles can be voiced in one call. The voice_id field offers voice_fish_narrator, and the result is a hosted MP3 delivered asynchronously via polling or webhook.

Fish Audio pricing on Muapi is metered on UTF-8 bytes of text: $15 per 1M bytes. English text is about one byte per character, so it works out to roughly $0.015 per 1,000 English characters; non-Latin scripts use three to four bytes per character and cost proportionally more per character.

Fish Audio API capabilities

Long-form input

Up to 10,000 characters per request, enough for scripts, chapters, and articles in one call.

MP3 output

Each job returns a hosted MP3 file you can embed, download, or pass to a video or lipsync pipeline.

Byte-based pricing

$15 per 1M UTF-8 bytes, about $0.015 per 1,000 English characters, with no subscription.

Async submit and poll

Use the same request ID, polling, and webhook flow as every other Muapi endpoint.

Fish Audio API use cases

Audiobooks and podcasts

Long-form narration for audiobooks and podcasts.

Explainers and courses

Voiceovers for explainer videos and course material.

Read-aloud

Read articles and documents aloud.

Apps and games

Multilingual narration for apps and games.

Fish Audio pricing and TTS model comparison

Model	Provider	Price	Best For
Fish Audio S2.1 Pro	Fish Audio	$15/1M UTF-8 bytes (~$0.015/1,000 English characters)	Long-form narration, up to 10,000 characters
Eleven v4	ElevenLabs	$0.08/1,000 characters	Expressive narration, up to 2,000 characters
Gemini 3.8 Flash TTS	Google	$0.015/1,000 characters	Multi-speaker dialogue
MiniMax Speech 2.6 Turbo	MiniMax	$0.07/1,000 characters	Fast single-voice generation

Fish Audio S2.1 Pro is billed per UTF-8 byte; non-Latin scripts use three to four bytes per character.

Fish Audio API example

Submit text to the Fish Audio S2.1 Pro API with your Muapi key, save the request ID, and poll the standard prediction result endpoint for the MP3 URL.

Generate speech with the Fish Audio API

curl -X POST https://api.muapi.ai/api/v1/fish-s2-1-pro \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"prompt":"Welcome to the show. Today we look at how voice models have changed the way we tell stories.","voice_id":"voice_fish_narrator"}'

# Response: {"request_id":"REQUEST_ID"}

fish-s2-1-pro takes prompt (up to 10,000 characters) and voice_id (voice_fish_narrator).

Poll the audio result

curl https://api.muapi.ai/api/v1/predictions/REQUEST_ID/result \
  -H "x-api-key: YOUR_API_KEY"

# Poll until status is completed, then read the generated MP3 URL from outputs.

Poll until status is completed, then read the MP3 URL from the response.

Fish Audio API FAQ

What is the Fish Audio API?

The Fish Audio API on Muapi is the fish-s2-1-pro endpoint, which runs Fish Audio S2.1 Pro for long-form text-to-speech and returns an MP3.

How much does the Fish Audio API cost?

Fish Audio pricing on Muapi is $15 per 1M UTF-8 bytes of text. English text is about one byte per character, so that is roughly $0.015 per 1,000 English characters. Non-Latin scripts use three to four bytes per character.

What is the Fish Audio S2.1 Pro API character limit?

Each request accepts up to 10,000 characters in prompt.

Is this the same as the Fish Speech API?

Fish Speech API is a name people also search for. This page covers the hosted Fish Audio S2.1 Pro model, called through Muapi's REST endpoint, so there is nothing to self-host. It exposes one voice, voice_fish_narrator.

What audio format does the Fish Audio text to speech API return?

A hosted MP3 file URL from the standard prediction result endpoint.

How do I call the Fish Audio S2.1 Pro API?

Create a Muapi API key, POST prompt and voice_id to /api/v1/fish-s2-1-pro, then poll /api/v1/predictions/{request_id}/result or pass a webhook URL.

Compare text to speech APIs →See Fish Audio S2.1 Pro beside MiniMax, Gemini, and ElevenLabs.Eleven v4 API →Expressive narration up to 2,000 characters per request.Gemini TTS API →Multi-speaker dialogue across four Gemini TTS tiers.Browse audio models →Compare speech, music, sound-effect, and video-to-audio endpoints.

Ready to use the Fish Audio API?

Voice long scripts with Fish Audio S2.1 Pro through one Muapi API key.

Get an API Key Try Fish Audio S2.1 Pro