aid: fish-audio
name: Fish Audio
description: >-
  Fish Audio is an AI voice platform offering text-to-speech, voice cloning,
  speech-to-text, voice changing, and audio storytelling capabilities. The
  platform hosts a library of over two million voices across 30+ languages and
  is built around the Fish Speech open-source TTS model and the proprietary
  Fish Audio S2-Pro model. Fish Audio exposes a public REST API at
  api.fish.audio with first-party Python, Go, and TypeScript SDKs and supports
  voice cloning from as little as fifteen seconds of reference audio. The
  developer surface emphasizes ultra-low latency streaming, emotion control,
  and pay-as-you-go pricing for both prototype and production workloads.
type: Index
position: Provider
access: 3rd-Party
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
tags:
  - Voice
  - Text to Speech
  - Speech to Text
  - Voice Cloning
  - Audio
  - Generative AI
  - Multilingual
  - Streaming
  - SDK
  - Open Source
url: https://raw.githubusercontent.com/api-evangelist/fish-audio/refs/heads/main/apis.yml
created: '2026-05-23'
modified: '2026-05-23'
specificationVersion: '0.20'
apis:
  - aid: fish-audio:fish-audio-api
    name: Fish Audio API
    description: >-
      The Fish Audio API provides RESTful access to text-to-speech,
      speech-to-text, voice cloning, and voice management capabilities backed
      by the Fish Audio S2-Pro model. Endpoints support streaming low-latency
      generation, multilingual synthesis across 30+ languages, emotion control,
      and on-the-fly custom voice creation from short reference clips. The API
      is consumed through the Fish Audio Python, Go, and TypeScript SDKs and a
      community of integrations including n8n.
    humanURL: https://docs.fish.audio
    baseURL: https://api.fish.audio
    tags:
      - Text to Speech
      - Voice Cloning
      - Speech to Text
      - Streaming
      - REST
      - Audio
    properties:
      - type: Documentation
        url: https://docs.fish.audio
      - type: GettingStarted
        url: https://docs.fish.audio/quickstart
      - type: Playground
        url: https://fish.audio/discovery
      - type: SDK
        url: https://github.com/fishaudio/fish-audio-python
      - type: SDK
        url: https://github.com/fishaudio/fish-audio-go
      - type: GitHubOrganization
        url: https://github.com/fishaudio
    features:
      - name: Text-to-Speech Generation
        description: >-
          Synthesize natural, emotionally expressive speech from text using
          the Fish Audio S2-Pro model across 30+ languages.
      - name: Voice Cloning
        description: >-
          Create custom voice models from as little as 15 seconds of reference
          audio for downstream TTS.
      - name: Speech-to-Text Transcription
        description: >-
          Transcribe audio with multispeaker detection and emotion tagging
          metadata.
      - name: Streaming Audio
        description: >-
          Low-latency streaming responses suitable for real-time agent, IVR,
          and live narration use cases.
      - name: Emotion and Prosody Control
        description: >-
          Inline emotion tags (angry, sad, excited) and special effects
          (laughing, sobbing) for expressive output.
      - name: Multilingual Synthesis
        description: >-
          Native support for English, Mandarin, Japanese, Korean, and more
          than 25 additional languages.
      - name: Voice Library
        description: >-
          Access to a hosted library of more than two million pre-built
          voices for instant TTS generation.
    useCases:
      - name: Audiobook and Podcast Production
        description: >-
          Generate full-length narrated content with multi-character voices
          via Story Studio workflows.
      - name: Conversational Agents and IVR
        description: >-
          Power voice-first agents and interactive voice response systems
          with low-latency synthesis.
      - name: Gaming NPC Dialogue
        description: >-
          Create dynamic in-game character voices and barks without manual
          voice-over sessions.
      - name: Video and Content Localization
        description: >-
          Dub and localize video, social, and marketing content across
          dozens of languages.
      - name: Accessibility Tooling
        description: >-
          Embed expressive screen reading and assistive voice output in
          accessibility products.
    integrations:
      - name: Python SDK
      - name: Go SDK
      - name: TypeScript SDK
      - name: n8n
      - name: LangChain
      - name: Hugging Face
      - name: Discord
    authentication:
      - type: API Key
        description: >-
          Requests authenticate using a Bearer API key issued from the Fish
          Audio dashboard.
common:
  - type: Website
    url: https://fish.audio
  - type: Documentation
    url: https://docs.fish.audio
  - type: Developer Portal
    url: https://fish.audio/go-api
  - type: Playground
    url: https://fish.audio/discovery
  - type: Pricing
    url: https://fish.audio/pricing
  - type: GitHubOrganization
    url: https://github.com/fishaudio
  - type: OpenSourceModel
    url: https://github.com/fishaudio/fish-speech
  - type: Discord
    url: https://discord.gg/Es5qTB9BcN
  - type: Twitter
    url: https://twitter.com/FishAudio
  - type: LLMsTxt
    url: https://docs.fish.audio/llms.txt
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com