Basic Information

appmixer.ai.deepgram.TranscribeAudio

Transcribe Audio

1.0.0

Public

Transcribes pre-recorded audio. Submits the job to Deepgram and returns the request ID immediately on the "out" port without holding a worker, then emits the finished transcript on the "done" port when Deepgram delivers it. Accepts an audio URL or a file (max 2 GB), with optional diarization, summarization, sentiment, topics and intents.

Input Ports

in

FieldTypeDescription
Audio URL text Publicly accessible URL of the audio file to transcribe. Provide either this or a File below.
File filepicker Audio file to transcribe (max 2 GB). Provide either this or an Audio URL above.
Model text Speech-to-text model, e.g. nova-3 (default). Start typing to pick from the available STT models, or enter a model name manually.
Language text BCP-47 language code, e.g. en. Leave empty to use the model default.
Detect Language toggle Automatically detect the spoken language.
Smart Format toggle Apply formatting to dates, times, numbers, etc.
Punctuate toggle Add punctuation and capitalization.
Diarize toggle Recognize and label distinct speakers.
Summarize toggle Generate a summary (Audio Intelligence).
Sentiment toggle Analyze sentiment (Audio Intelligence).
Topics toggle Detect topics (Audio Intelligence).
Intents toggle Detect intents (Audio Intelligence).
Additional Query Parameters key-value Any additional Deepgram query parameters (e.g. redact, keyterm, search, replace).
Correlation ID text Optional. Any value of your own — an order number, a file name, a record ID. It is echoed back on both the out and done ports, so when several transcriptions run at once you can tell which finished transcript belongs to which input.

Output Ports

out

FieldTypeDescription
Request ID string -
Audio URL string -
File ID string -
Correlation ID string -

done

FieldTypeDescription
Request ID string -
Audio URL string -
File ID string -
Correlation ID string -
Error string -
Transcript string -
Metadata object -
Results object -

Full Metadata