GASTON offers a simple REST API interface that allows you to connect your applications and websites to our services and transcribe videos and audio tracks.
The actual file transfer is asynchronous, so after uploading the file to our API, you will receive
ID
file, which is processed sequentially. The audio track is extracted, a spectrogram is created, the language is detected, and the transcription itself is performed. The current state of the file,
as well as the acquisition of the transcript and other information, is carried out via the endpoint /media.
All overwritten files are saved into library,
just as if they were added through our web app.
The API key must be inserted into the HTTP Header with each request. After logging in your settings generate an API key that is used for authorization against our REST API. For example:
curl -X 'GET'
'https://api.gaston.live/user/me'
-H 'accept: application/json'
-H 'token: gapi-1234567890'
You can change your key at any time. in your settings.
Exact technical specification for the openapi 3.1 format is here, a simple description of the endpoints follows.
The endpoint is used to obtain basic information about the authenticated user and also contains information about the service usage - monthly quota and current service usage.
The endpoint returns complete information about the file, including transcription, word timing, and available translations.
Except for the parameter media_id parameter is also available lang which
defines which language mutation of the transcript should be listed. If specified lang=en and the file is translated into English,
the English sentences are listed.
Endpoint for listing already processed files in the library. The only parameter available is the page number you want to list. Files are paged through 50 records and are sorted from newest to oldest. The information about the files is brief and does not contain the actual text transcript of the file.
Endpoint for uploading a media file for transcription. Maximum file size is 2Gb. The endpoint has the ability
to specify the language spoken in the recording.
If the language is not specified, it will be automatically detected. If the language is known, it is recommended
to specify it for better speed. The endpoint also supports
setting the file name in the library. If not specified, the file name will be used. After a successful call
the endpoint will respond with information about ID file and its status.
Like the previous endpoint, this one is used to transcribe a video from a URL address of one of the supported streaming services. After entering the video URL, the file will be automatically downloaded and processed. For example, if the URL address is provided https://www.youtube.com/watch?v=-fke4tqK0cY, GASTON will download the video and all the information about it, including the image, and of course overwrite the file.
Translates an already transcribed file into the selected language. Currently, only English is available. The progress
of the translation into the given language,
as well as the available language mutations for each file, is indicated by an associative fieldavailable_languages.
Where the key is the language name and the value is the percentage of translation progress.
For example, the following status means that the file is 100% transcribed into Slovak, and the translation into English is still in progress and is at 66%:
available_languages: {
sk: 100,
en: 66
}
Our API is not limited in any way (except for the monthly quota according to the user's plan). For technical support, write to us at any time at contact@streams.guru. We are constantly expanding endpoints and functionality.