Listen — From audio file asynchronously
POST/listen_async
Generate a transcript and a structured clinical note from an audio file. Only audio/*
mime types are supported. The maximum duration is 60 minutes. The result of the request is posted to the callback URL once available.
Request
- application/json
Body
required
The URL of the audio file to download & listen to.
Possible values: [transcript_item
, note
]
Specifies which items you want us to send you back. In other words, which feature(s) you want to use, transcription and/or note generation.
Possible values: [fr
, en
, en-US
, en-GB
, fr-FR
, es-ES
, es-MX
]
Language spoken in the audio ('fr' and 'en' are deprecated, and correspond to 'fr-FR' and 'en-US' respectively).
Default value: false
Indicates whether to segment transcription results at sentence boundaries. Default is false, meaning that a single transcript item may encompass multiple sentences, provided they are not delineated by pauses (silence) in the audio.
Possible values: [highest_quality
, fastest
]
Choose a generation mode:
• highest_quality
: generates very high quality notes, might take up to one minute;
• fastest
: quicker note generation (few seconds), but might not give the best possible output.
Default is highest_quality
.
Possible values: [GENERAL_MEDICINE
, CARDIOLOGY
, PSYCHIATRY
, DIET
, PSYCHOLOGY
, SOAP
]
The desired template of the generated note. Default is GENERAL_MEDICINE
.
Check Note template for details.
Once we are done processing the input and generating the transcript & note, we will call you back with a 'POST' on this URL and a Json body respecting exactly the same schema as the response of the synchronous equivalent endpoint (i.e. /copilot/listen
).
Possible values: <= 128 characters
An optional unique client-made id for this request. When specified, this id is included in the Json payload that will be sent asynchronously. This can help clients both guarantee idempotency and correlate asynchronous responses with their requests. Note: this id must be different for every request. It is recommended to use a randomly generated UUID.
Responses
- 200
- 299
Request received and processing started.
- application/json
- Schema
- Example (from schema)
Schema
The id of the request you just created.
The provided client_request_id
.
{
"request_id": "98FCE1EF-DBCA-41EF-8BC7-4D1621AC07C6",
"client_request_id": "string"
}
Here's the schema of the Json payload that will be sent asynchronously to the URL you provided. (299 is not a possible response code — only a way here to document the asynchronous payload schema).
- application/json
- Schema
Schema
- success
- error
- Array [
- ]
- Array [
- ]
The id of the async digest request.
The client_request_id
provided in the request.
transcript object[]
Transcript items from the audio file.
The transcribed text.
Possible values: [doctor
, patient
, unspecified
]
Who said the text in this transcript item.
Start time of this transcription item as the offset, in milliseconds, from the start of the audio file.
End time of this transcription item as the offset, in milliseconds, from the start of the audio file. Equals the start_time_ms
plus the duration of the related transcribed audio portion.
note object
The generated note.
sections object[]required
Content of the note structured in multiple sections.
A key identifying a section of a note. The set of possible keys depend on the template that is used. Check Note template for possible values.
The section title.
Content of the note section.
The id of the async digest request.
The client_request_id
provided in the request.