Audio Transcriber for Audio and Video URLs · ElevenLabs

{
  "type": "object",
  "required": [
    "body"
  ],
  "properties": {
    "body": {
      "type": "object",
      "$schema": "https://json-schema.org/draft/2020-12/schema",
      "required": [
        "audio_url"
      ],
      "~standard": {
        "vendor": "zod",
        "version": 1,
        "jsonSchema": {}
      },
      "properties": {
        "diarize": {
          "type": "boolean",
          "description": "Annotate which speaker is talking (up to 32 speakers)."
        },
        "keyterms": {
          "type": "array",
          "items": {
            "type": "string",
            "maxLength": 50,
            "minLength": 1
          },
          "maxItems": 100,
          "description": "Up to 100 domain terms (product names, jargon) to bias transcription toward. ADD-ON: raises the rate from $0.22 to $0.27 per hour of audio."
        },
        "model_id": {
          "enum": [
            "scribe_v2",
            "scribe_v1"
          ],
          "type": "string",
          "default": "scribe_v2",
          "description": "Transcription model: scribe_v2 (state of the art, 90+ languages) or scribe_v1 (legacy)."
        },
        "audio_url": {
          "type": "string",
          "format": "uri",
          "description": "HTTPS URL of the audio or video file to transcribe (public or presigned link; hosted media, YouTube/TikTok URLs also work). ElevenLabs fetches it directly — max 2GB."
        },
        "no_verbatim": {
          "type": "boolean",
          "description": "Remove filler words, false starts and disfluencies (scribe_v2 only)."
        },
        "num_speakers": {
          "type": "integer",
          "maximum": 32,
          "minimum": 1,
          "description": "Maximum number of speakers in the file (helps accuracy)."
        },
        "language_code": {
          "type": "string",
          "maxLength": 8,
          "minLength": 2,
          "description": "Optional ISO-639-1/-3 language code of the audio. Omit to auto-detect."
        },
        "entity_detection": {
          "type": "boolean",
          "description": "Detect entities (names, credit cards, SSNs, medical conditions, …) with exact positions. ADD-ON: raises the rate from $0.22 to $0.29 per hour of audio."
        },
        "tag_audio_events": {
          "type": "boolean",
          "description": "Tag non-speech audio events like (laughter). Default true."
        },
        "timestamps_granularity": {
          "enum": [
            "none",
            "word",
            "character"
          ],
          "type": "string",
          "description": "Timestamp granularity (default word)."
        },
        "expected_duration_minutes": {
          "type": "number",
          "maximum": 600,
          "description": "Monid-only billing hint (NOT sent upstream): expected audio length in minutes, used to size the admission hold. Settlement always bills the actual transcribed duration. Omitted → a conservative 60-minute hold.",
          "exclusiveMinimum": 0
        }
      },
      "description": "Transcribe an audio or video file from a URL",
      "additionalProperties": false
    }
  },
  "additionalProperties": false
}