For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Deepgram has released an update to our Speech Recognition API. Improvements affect the following areas:
Keyword Boosting for Pre-Recorded Transcription
We have increased the accuracy of out-of-vocabulary (OOV) keyword boosting by reducing false positives when processing pre-recorded audio.
Transcription Response Objects
Response objects sent from transcription endpoints (/listen) will now include unique identifiers for models:
Batch: The metadata object will now include a models field (array of strings), which lists the unique identifiers of all models used during transcription. View a sample response object.
Streaming: For each interim transcript, the metadata object will now include a model_uuid field (string), which lists the unique identifier for the model used to transcribe the particular channel and chunk of data. The last response returned during streaming will not include this field. View a sample response object, or to learn more about interim transcripts, see Understand Interim Transcripts in Deepgram’s Developer Documentation.
Transcript: The metadata object will now include a models field (array of strings), which includes the unique identifiers of all models used during transcription. View a sample response object.
Deepgram support is always available to walk customers through updates, as well as resolve any issues that arise when upgrading.
