Features & formats

Every request can enable the full feature set — there are no paid add-ons.

Transcription options

Option	Default	Description
`word_timestamps`	`true`	Start and end time of each word, in seconds
`speaker_labels`	`true`	Speaker diarization (who spoke when)
`diarize`	—	Deepgram-compatible alias for `speaker_labels`
`nltk`	`true`	Restore punctuation & capitalization
`custom_vocabulary`	—	Names and terms to recognize
`language`	auto	ISO 639-1 code (e.g. `en`) to skip detection and transcribe the whole file in that language — see pinning the language
`tier`	`standard`	`standard` — the only tier currently available
`output_type`	`json`	Result file format

Output types

Type	Use for
`json`	Structured `words`, plus diarization when enabled. The SDKs parse this into `text`, `words`, and `utterances`.
`txt`	Plain transcript, in speaker blocks when `speaker_labels` is on
`srt` / `vtt`	Subtitles / captions
`docx` / `pdf`	Shareable documents

JSON shape (native)

The default JSON output looks like this:

{
  "words": [
    {"word": "Hello,", "start": 0.12, "end": 0.40, "speaker": "SPEAKER_1", "language": "en"},
    {"word": "world.", "start": 0.41, "end": 0.70, "speaker": "SPEAKER_1", "language": "en"}
  ],
  "diarization": [
    {"start": 0.12, "end": 0.70, "speaker": "SPEAKER_1"}
  ],
  "languages": [
    {"start": 0.12, "end": 0.70, "language": "en"}
  ]
}

Live progress

Every SDK reports progress for both the file upload and the transcription, either as console bars (the progress option) or through onProgress / onUploadProgress callbacks. Each callback receives completed, total, step, and a computed percent. The values come from the job SSE stream's completed/total counts, so they track completed work, not an estimate.

SDK adapters

The Python and JavaScript SDKs expose AssemblyAI-style result.text and result.utterances, plus result.to_deepgram() for code that expects Deepgram's response shape.