aid: neuphonic
name: Neuphonic
type: Index
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
url: https://raw.githubusercontent.com/api-evangelist/neuphonic/refs/heads/main/apis.yml
created: 2026-06-12
modified: 2026-06-12
specificationVersion: '0.19'
description: >
  Neuphonic is an ultra-low-latency voice AI platform specializing in real-time
  text-to-speech synthesis with sub-25ms latency, making it suitable for
  conversational AI and live applications. The platform provides both a
  cloud-hosted API with WebSocket streaming and Server-Sent Events (SSE), as
  well as open-source on-device models (NeuTTS Air, NeuTTS Nano) that run
  without a GPU. Neuphonic supports nine languages including English, Spanish,
  German, French, Urdu, Japanese, Korean, Chinese, and Portuguese, and offers
  instant voice cloning from short audio samples. Developers can also build
  conversational AI agents via the Agent API, which integrates with GPT-4o and
  supports Model Context Protocol (MCP) servers. Authentication uses API keys
  passed via the X-API-KEY header for SSE and as a query parameter for
  WebSocket connections.
tags:
  - Text-to-Speech
  - Voice AI
  - Audio
  - Streaming
  - WebSocket
  - Voice Cloning
  - Conversational AI
  - Real-Time
  - Multilingual
  - On-Device AI
apis:
  - name: Neuphonic TTS SSE API
    description: >
      Server-Sent Events endpoint for real-time text-to-speech synthesis.
      Accepts POST requests with text content and returns streaming audio in
      PCM format. Supports language selection, voice selection, speed control,
      and sampling rate configuration.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.neuphonic.com/build-group/text-to-speech
    baseURL: https://api.neuphonic.com
    tags:
      - Text-to-Speech
      - SSE
      - Streaming
      - Audio
    properties:
      - type: Documentation
        url: https://docs.neuphonic.com/build-group/text-to-speech
      - type: OpenAPI
        url: https://docs.neuphonic.com/api-reference/tts/sse
      - type: OpenAPI
        url: openapi/neuphonic-openapi.yml

  - name: Neuphonic TTS WebSocket API
    description: >
      WebSocket endpoint for continuous, low-latency text-to-speech streaming.
      Enables real-time voice synthesis with sub-25ms latency, supporting
      multiple text chunks over a single persistent connection. Includes a
      60-second connection timeout and requires a STOP token to signal
      completion.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.neuphonic.com/build-group/text-to-speech
    baseURL: wss://api.neuphonic.com
    tags:
      - Text-to-Speech
      - WebSocket
      - Real-Time
      - Streaming
    properties:
      - type: Documentation
        url: https://docs.neuphonic.com/build-group/text-to-speech
      - type: AsyncAPI
        url: https://docs.neuphonic.com/api-reference/tts/websocket

  - name: Neuphonic Voice Cloning API
    description: >
      REST API for creating and managing cloned voices. Accepts audio samples
      (MP3 or WAV, minimum 6 seconds, under 10MB) and generates a custom voice
      model. Supports creating, retrieving, updating, listing, and deleting
      cloned voices. Voice cloning is available in English, French, German, and
      Spanish.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.neuphonic.com/build-group/voice-cloning
    baseURL: https://api.neuphonic.com
    tags:
      - Voice Cloning
      - Audio
      - Machine Learning
    properties:
      - type: Documentation
        url: https://docs.neuphonic.com/build-group/voice-cloning

  - name: Neuphonic Agent API
    description: >
      REST API for creating and managing conversational AI voice agents. Agents
      combine Neuphonic TTS with GPT-4o for interactive voice applications and
      support Model Context Protocol (MCP) server integration, multilingual
      operation, and interruption handling for natural conversational flow.
    image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
    humanURL: https://docs.neuphonic.com/
    baseURL: https://api.neuphonic.com
    tags:
      - Agents
      - Conversational AI
      - Voice Assistant
    properties:
      - type: Documentation
        url: https://docs.neuphonic.com/

common:
  - type: Website
    url: https://www.neuphonic.com/
  - type: Documentation
    url: https://docs.neuphonic.com/
  - type: GitHubOrganization
    url: https://github.com/neuphonic
  - type: LinkedIn
    url: https://uk.linkedin.com/company/neuphonic
  - type: X
    url: https://twitter.com/neuphonicspeech
  - type: StatusPage
    url: https://status.neuphonic.com/
  - type: Playground
    url: https://app.neuphonic.com/
  - type: Plans
    url: plans/neuphonic-plans-pricing.yml
  - type: RateLimits
    url: rate-limits/neuphonic-rate-limits.yml
  - type: FinOps
    url: finops/neuphonic-finops.yml
  - type: Vocabulary
    url: vocabulary/neuphonic-vocabulary.yml
  - type: JSONLDContext
    url: json-ld/neuphonic-context.jsonld

maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com