2025-07-05 140856,900 - ERROR livekit.agents - unhandled exception while running the job task Traceback (most recent call last): File "/Users/sandeep/projects/line/agent/outbound/handler.py", line 487, in entrypoint stt=deepgram.STT(api_key=DEEPGRAM_API_KEY, language="multi", File "/Users/sandeep/projects/agntv1/lib/python3.10/site-packages/livekit/plugins/deepgram/stt.py", line 141, in init _validate_keyterms(model, language, keyterms, keywords) File "/Users/sandeep/projects/agntv1/lib/python3.10/site-packages/livekit/plugins/deepgram/stt.py", line 729, in _validate_keyterms raise ValueError( ValueError: Keyterm Prompting is only available for English transcription using the Nova-3 Model. To boost recognition of keywords using another model, use the Keywords feature. {"pid": 98203, "job_id": "AJ_7xMtViWnzeph"}

Thanks for reaching out, Sandeep. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

I am using language as multi

Hey there! I totally get the confusion with language settings. Just wanted to let you know that the "multi" language option won't work with Keyterm Prompting – that's only available for English transcriptions using the Nova-3 model. But don't worry! If you want to help boost recognition of specific words while using the "multi" language setting, the Keywords feature is your go-to solution. Hope this helps clear things up!

but is Keywords available in nova-3-general

Hey there! I totally get how this might be confusing. Just wanted to let you know that the Keywords feature isn't supported in the Nova-3-General model. If you're working with Nova-3, no worries – you can use Keyterm Prompting instead. For Keywords, you'll want to stick with Nova-2, Nova-1, Enhanced, or Base speech-to-text models. Hope this helps clear things up! Sources: GitHub::livekit::agents | livekit-plugins/livekit-plugins-deepgram/livekit/...

how would this keyterm prompting look like in code

Keyterm is available only for english in deepgram