When working with Deepgram's transcription services, users often seek to include filler words such as "um," "uh," or "mm" in their transcriptions. This functionality is particularly useful for creating more natural transcriptions that capture the nuances of spoken language.
To enable the transcription of filler words in Deepgram's services, you need to set the filler_words=true
parameter in your API request. This request can be applied to both pre-recorded and streaming audio, but it's important to note that this feature currently supports English language audio only.
For those using Deepgram's API, ensure your request includes the correct parameter:
https://api.deepgram.com/v1/listen
for regular transcription or wss://api.deepgram.com/v1/listen
for live transcription.filler_words=true
to your request to enable this feature.More detailed documentation on setting this parameter and its use can be found in our filler words documentation.
Despite enabling filler words, there can be situations where some filler words are not captured accurately:
By appropriately setting up your API request and ensuring high-quality input audio, you can fully utilize Deepgram's ability to detect and transcribe filler words. If issues persist or the system behavior seems inconsistent, reach out to your Deepgram support representative (if you have one) or visit our community for assistance: https://discord.gg/deepgram