Together AI · Capability

Together APIs — Audio

Together APIs — Audio. 5 operations. Lead operation: Create audio generation request. Self-contained Naftiko capability covering one Together Ai business surface.

Run with Naftiko Together AiAudio

What You Can Do

POST
Audiospeech — Create audio generation request
/v1/audio/speech
GET
Realtimetts — Real-time text-to-speech via WebSocket
/v1/audio/speech/websocket
POST
Audiotranscriptions — Create audio transcription request
/v1/audio/transcriptions
POST
Audiotranslations — Create audio translation request
/v1/audio/translations
GET
Realtimetranscription — Real-time audio transcription via WebSocket
/v1/realtime

MCP Tools

create-audio-generation-request

Create audio generation request

real-time-text-speech-websocket

Real-time text-to-speech via WebSocket

read-only idempotent
create-audio-transcription-request

Create audio transcription request

create-audio-translation-request

Create audio translation request

real-time-audio-transcription-websocket

Real-time audio transcription via WebSocket

read-only idempotent

Capability Spec

together-ai-audio.yaml Raw ↑
naftiko: 1.0.0-alpha2
info:
  label: Together APIs — Audio
  description: 'Together APIs — Audio. 5 operations. Lead operation: Create audio generation request. Self-contained Naftiko
    capability covering one Together Ai business surface.'
  tags:
  - Together Ai
  - Audio
  created: '2026-05-19'
  modified: '2026-05-19'
binds:
- namespace: env
  keys:
    TOGETHER_AI_API_KEY: TOGETHER_AI_API_KEY
capability:
  consumes:
  - type: http
    namespace: together-ai-audio
    baseUri: https://api.together.ai/v1
    description: Together APIs — Audio business capability. Self-contained, no shared references.
    resources:
    - name: audio-speech
      path: /audio/speech
      operations:
      - name: audiospeech
        method: POST
        description: Create audio generation request
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
    - name: audio-speech-websocket
      path: /audio/speech/websocket
      operations:
      - name: realtimetts
        method: GET
        description: Real-time text-to-speech via WebSocket
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: model
          in: query
          type: string
        - name: voice
          in: query
          type: string
        - name: max_partial_length
          in: query
          type: integer
        - name: language
          in: query
          type: string
    - name: audio-transcriptions
      path: /audio/transcriptions
      operations:
      - name: audiotranscriptions
        method: POST
        description: Create audio transcription request
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: true
    - name: audio-translations
      path: /audio/translations
      operations:
      - name: audiotranslations
        method: POST
        description: Create audio translation request
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: true
    - name: realtime
      path: /realtime
      operations:
      - name: realtimetranscription
        method: GET
        description: Real-time audio transcription via WebSocket
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: model
          in: query
          type: string
          required: true
        - name: input_audio_format
          in: query
          type: string
          description: Audio format specification. Currently supports 16-bit PCM at 16kHz sample rate.
          required: true
    authentication:
      type: bearer
      token: '{{env.TOGETHER_AI_API_KEY}}'
  exposes:
  - type: rest
    namespace: together-ai-audio-rest
    port: 8080
    description: REST adapter for Together APIs — Audio. One Spectral-compliant resource per consumed operation, prefixed
      with /v1.
    resources:
    - path: /v1/audio/speech
      name: audio-speech
      description: REST surface for audio-speech.
      operations:
      - method: POST
        name: audiospeech
        description: Create audio generation request
        call: together-ai-audio.audiospeech
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/audio/speech/websocket
      name: audio-speech-websocket
      description: REST surface for audio-speech-websocket.
      operations:
      - method: GET
        name: realtimetts
        description: Real-time text-to-speech via WebSocket
        call: together-ai-audio.realtimetts
        with:
          model: rest.model
          voice: rest.voice
          max_partial_length: rest.max_partial_length
          language: rest.language
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/audio/transcriptions
      name: audio-transcriptions
      description: REST surface for audio-transcriptions.
      operations:
      - method: POST
        name: audiotranscriptions
        description: Create audio transcription request
        call: together-ai-audio.audiotranscriptions
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/audio/translations
      name: audio-translations
      description: REST surface for audio-translations.
      operations:
      - method: POST
        name: audiotranslations
        description: Create audio translation request
        call: together-ai-audio.audiotranslations
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/realtime
      name: realtime
      description: REST surface for realtime.
      operations:
      - method: GET
        name: realtimetranscription
        description: Real-time audio transcription via WebSocket
        call: together-ai-audio.realtimetranscription
        with:
          model: rest.model
          input_audio_format: rest.input_audio_format
        outputParameters:
        - type: object
          mapping: $.
  - type: mcp
    namespace: together-ai-audio-mcp
    port: 9090
    transport: http
    description: MCP adapter for Together APIs — Audio. One tool per consumed operation, routed inline through this capability's
      consumes block.
    tools:
    - name: create-audio-generation-request
      description: Create audio generation request
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: together-ai-audio.audiospeech
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: real-time-text-speech-websocket
      description: Real-time text-to-speech via WebSocket
      hints:
        readOnly: true
        destructive: false
        idempotent: true
      call: together-ai-audio.realtimetts
      with:
        model: tools.model
        voice: tools.voice
        max_partial_length: tools.max_partial_length
        language: tools.language
      outputParameters:
      - type: object
        mapping: $.
    - name: create-audio-transcription-request
      description: Create audio transcription request
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: together-ai-audio.audiotranscriptions
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: create-audio-translation-request
      description: Create audio translation request
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: together-ai-audio.audiotranslations
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: real-time-audio-transcription-websocket
      description: Real-time audio transcription via WebSocket
      hints:
        readOnly: true
        destructive: false
        idempotent: true
      call: together-ai-audio.realtimetranscription
      with:
        model: tools.model
        input_audio_format: tools.input_audio_format
      outputParameters:
      - type: object
        mapping: $.