{
  "name": "Text to Speech (OpenAI)",
  "tags": [],
  "nodes": [
    {
      "id": "938fedbd-e34c-40af-af2f-b9c669e1a6e9",
      "name": "When clicking \"Test workflow\"",
      "type": "n8n-nodes-base.manualTrigger",
      "position": [
        380,
        380
      ],
      "parameters": {},
      "typeVersion": 1
    },
    {
      "id": "1d59db5d-8fe6-4292-a221-a0d0194c6e0c",
      "name": "Set input text and TTS voice",
      "type": "n8n-nodes-base.set",
      "position": [
        760,
        380
      ],
      "parameters": {
        "assignments": {
          "assignments": [
            {
              "id": "speech-text",
              "name": "input_text",
              "type": "string",
              "value": "The quick brown fox jumped over the lazy dog."
            },
            {
              "id": "speech-voice",
              "name": "voice",
              "type": "string",
              "value": "alloy"
            }
          ]
        },
        "options": {}
      },
      "typeVersion": 3.4
    },
    {
      "id": "9d54de1d-59b7-4c1f-9e88-13572da5292c",
      "name": "Send HTTP Request to OpenAI's TTS Endpoint",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        1120,
        380
      ],
      "parameters": {
        "url": "https://api.openai.com/v1/audio/speech",
        "method": "POST",
        "options": {
          "response": {
            "response": {
              "responseFormat": "file",
              "outputPropertyName": "data"
            }
          }
        },
        "sendBody": true,
        "sendHeaders": false,
        "authentication": "predefinedCredentialType",
        "bodyParameters": {
          "parameters": [
            {
              "name": "model",
              "value": "tts-1"
            },
            {
              "name": "input",
              "value": "={{ $json.input_text }}"
            },
            {
              "name": "voice",
              "value": "={{ $json.voice }}"
            },
            {
              "name": "response_format",
              "value": "mp3"
            }
          ]
        },
        "nodeCredentialType": "openAiApi"
      },
      "typeVersion": 4.2
    },
    {
      "id": "1ce72c9c-aa6f-4a18-9d5a-3971686a51ec",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        280,
        256
      ],
      "parameters": {
        "width": 273,
        "height": 339,
        "content": "## Workflow Trigger\nYou can replace this manual trigger with another trigger type as required by your use case."
      },
      "typeVersion": 1
    },
    {
      "id": "eb487535-5f36-465e-aeee-e9ff62373e53",
      "name": "Sticky Note1",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        660,
        257
      ],
      "parameters": {
        "width": 273,
        "height": 335,
        "content": "## Manually Set OpenAI TTS Configuration\n"
      },
      "typeVersion": 1
    },
    {
      "id": "36b380bd-0703-4b60-83cb-c4ad9265864d",
      "name": "Sticky Note2",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1020,
        260
      ],
      "parameters": {
        "width": 302,
        "height": 335,
        "content": "## Send Request to OpenAI TTS API\n"
      },
      "typeVersion": 1
    },
    {
      "id": "ff35ff28-62b5-49c8-a657-795aa916b524",
      "name": "Sticky Note3",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        660,
        620
      ],
      "parameters": {
        "color": 4,
        "width": 273,
        "height": 278,
        "content": "### Input\n\nSet input_text to at most 4096 characters and voice to a supported tts-1 voice, initially alloy. Keep your OpenAI key in the credential. The text is sent to OpenAI for speech generation."
      },
      "typeVersion": 1
    },
    {
      "id": "5f7ef80e-b5c8-41df-9411-525fafc2d910",
      "name": "Sticky Note4",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1020,
        620
      ],
      "parameters": {
        "color": 4,
        "width": 299,
        "height": 278,
        "content": "### Output and availability\n\nThe HTTP node returns an MP3 in binary field data. No automatic retries are configured. Disclose that the recording is AI-generated when sharing it. OpenAI has announced tts-1 shutdown on 2027-01-06; plan migration before then. https://developers.openai.com/api/docs/guides/text-to-speech"
      },
      "typeVersion": 1
    },
    {
      "id": "validate-speech-input",
      "name": "Validate Speech Input",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        940,
        380
      ],
      "parameters": {
        "mode": "runOnceForAllItems",
        "jsCode": "const prepare = function prepareSpeech({ input_text, voice }) {\n  if (typeof input_text !== 'string' || !input_text.trim()) throw new Error('Input text is required.');\n  if (Array.from(input_text).length > 4096) throw new Error('Input text must contain at most 4096 characters.');\n  const selected = voice == null || voice === '' ? 'alloy' : voice;\n  if (!['alloy', 'ash', 'coral', 'echo', 'fable', 'onyx', 'nova', 'sage', 'shimmer'].includes(selected)) throw new Error('Choose a supported tts-1 voice.');\n  return { input_text, voice: selected };\n}\n\nreturn $input.all().map((item,index)=>({json:prepare(item.json),pairedItem:{item:index}}));\n"
      }
    },
    {
      "id": "validate-speech-file",
      "name": "Validate Audio File",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        1320,
        380
      ],
      "parameters": {
        "mode": "runOnceForAllItems",
        "jsCode": "const items = $input.all();\nconst result = [];\nfor (let index = 0; index < items.length; index++) {\n  const item = items[index];\n  const audio = item.binary?.data;\n  const type = typeof audio?.mimeType === 'string' ? audio.mimeType.split(';')[0].trim().toLowerCase() : '';\n  if (!['audio/mpeg', 'audio/mp3', 'application/octet-stream'].includes(type)) throw new Error('The speech response is not MP3 audio.');\n  const bytes = await this.helpers.getBinaryDataBuffer(index, 'data');\n  if (!bytes.length) throw new Error('The speech response is empty.');\n  result.push({...item, binary:{...item.binary,data:{...audio,fileName:'speech.mp3',mimeType:'audio/mpeg'}},pairedItem:{item:index}});\n}\nreturn result;\n"
      }
    }
  ],
  "active": false,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "connections": {
    "Set input text and TTS voice": {
      "main": [
        [
          {
            "node": "Validate Speech Input",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "When clicking \"Test workflow\"": {
      "main": [
        [
          {
            "node": "Set input text and TTS voice",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Validate Speech Input": {
      "main": [
        [
          {
            "node": "Send HTTP Request to OpenAI's TTS Endpoint",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Send HTTP Request to OpenAI's TTS Endpoint": {
      "main": [
        [
          {
            "node": "Validate Audio File",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}