vm6eji6m4/whisper-chinese-pro-fast

Premium L40S edition - 3x faster than T4. Mandarin/Taiwanese/Hakka podcast transcription with speaker diarization.

Public

1 run

GitHub

Run vm6eji6m4/whisper-chinese-pro-fast with an API

Use one of our client libraries to get started quickly. Clicking on a library will take you to the Playground tab where you can tweak different inputs, see the results, and copy the corresponding code to use in your own project.

Input schema

The fields you can use to run this model with an API. If you don't give a value for a field its default value will be used.

Field	Type	Default value	Description
audio	string		Audio file (mp3/wav/m4a/mp4). Or use file_url / file_string.
file_url	string		Audio file URL (alternative to `audio`). Public HTTPS URL.
file_string	string		Base64-encoded audio (alternative).
language	None		None
num_speakers	integer	0 Max: 10	Number of speakers (1-10). 0 = auto-detect.
prompt	string		None
use_builtin_vocab	boolean	True	None
enable_diarization	boolean	True	Run speaker diarization (requires hf_token).
gap_threshold	number	1.5 Min: 0.1 Max: 5	Merge adjacent same-speaker segments within this gap (seconds).
word_timestamps	boolean	False	Include per-word timestamps and per-word probability.
output_format	None	srt	Primary output format (JSON segments always included).
hf_token	string		None

{
  "type": "object",
  "title": "Input",
  "properties": {
    "audio": {
      "type": "string",
      "title": "Audio",
      "format": "uri",
      "x-order": 0,
      "nullable": true,
      "description": "Audio file (mp3/wav/m4a/mp4). Or use file_url / file_string."
    },
    "prompt": {
      "type": "string",
      "title": "Prompt",
      "default": "",
      "x-order": 5
    },
    "file_url": {
      "type": "string",
      "title": "File Url",
      "default": "",
      "x-order": 1,
      "description": "Audio file URL (alternative to `audio`). Public HTTPS URL."
    },
    "hf_token": {
      "type": "string",
      "title": "Hf Token",
      "format": "password",
      "x-order": 11,
      "nullable": true,
      "writeOnly": true,
      "x-cog-secret": true
    },
    "language": {
      "enum": [
        "",
        "zh",
        "zh-tw",
        "zh-hk",
        "en",
        "ja",
        "ko"
      ],
      "type": "string",
      "title": "language",
      "description": "An enumeration.",
      "default": "",
      "x-order": 3
    },
    "file_string": {
      "type": "string",
      "title": "File String",
      "default": "",
      "x-order": 2,
      "description": "Base64-encoded audio (alternative)."
    },
    "num_speakers": {
      "type": "integer",
      "title": "Num Speakers",
      "default": 0,
      "maximum": 10,
      "minimum": 0,
      "x-order": 4,
      "description": "Number of speakers (1-10). 0 = auto-detect."
    },
    "gap_threshold": {
      "type": "number",
      "title": "Gap Threshold",
      "default": 1.5,
      "maximum": 5,
      "minimum": 0.1,
      "x-order": 8,
      "description": "Merge adjacent same-speaker segments within this gap (seconds)."
    },
    "output_format": {
      "enum": [
        "srt",
        "vtt",
        "json",
        "plain"
      ],
      "type": "string",
      "title": "output_format",
      "description": "Primary output format (JSON segments always included).",
      "default": "srt",
      "x-order": 10
    },
    "word_timestamps": {
      "type": "boolean",
      "title": "Word Timestamps",
      "default": false,
      "x-order": 9,
      "description": "Include per-word timestamps and per-word probability."
    },
    "use_builtin_vocab": {
      "type": "boolean",
      "title": "Use Builtin Vocab",
      "default": true,
      "x-order": 6
    },
    "enable_diarization": {
      "type": "boolean",
      "title": "Enable Diarization",
      "default": true,
      "x-order": 7,
      "description": "Run speaker diarization (requires hf_token)."
    }
  }
}

Output schema

The shape of the response you’ll get when you run this model with an API.

Schema

{
  "type": "object",
  "title": "Output"
}