[GH-ISSUE #13720] feat: Support choosing a language while using the Deepgram Voice API. #55673

New Issue

GiteaMirror · 2026-05-05T17:48:55-05:00

GiteaMirror commented

2026-05-05 17:48:55 -05:00

Originally created by @homjay on GitHub (May 9, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/13720

Check Existing Issues

I have searched the existing issues and discussions.

Problem Description

When using the Deepgram API, I cannot choose a language. The default language of nova-3 is English, even when I speak other languages.

Desired Solution you'd like

Add a feature to select a language, using the official API example:


curl \
  -X POST \
  -H "Authorization: Token api" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://static.deepgram.com/examples/Bueller-Life-moves-pretty-fast.wav"}' \
  "https://api.deepgram.com/v1/listen?smart_format=true&language=multi&model=nova-3"

Alternatives Considered

No response

Additional Context

No response

Originally created by @homjay on GitHub (May 9, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/13720 ### Check Existing Issues - [x] I have searched the existing issues and discussions. ### Problem Description When using the Deepgram API, I cannot choose a language. The default language of nova-3 is English, even when I speak other languages. ### Desired Solution you'd like Add a feature to select a language, using the official API example: ``` curl \ -X POST \ -H "Authorization: Token api" \ -H "Content-Type: application/json" \ -d '{"url":"https://static.deepgram.com/examples/Bueller-Life-moves-pretty-fast.wav"}' \ "https://api.deepgram.com/v1/listen?smart_format=true&language=multi&model=nova-3" ``` ### Alternatives Considered _No response_ ### Additional Context _No response_

GiteaMirror closed this issue

2026-05-05 17:49:01 -05:00

GiteaMirror commented

2026-05-05 17:49:07 -05:00

@morbificagent commented on GitHub (May 9, 2025):

Yes i have tried using Deepgram nova-3 here too (germany) and noticed this too...
Would be great so we could use it in other languages

@morbificagent commented on GitHub (May 9, 2025): Yes i have tried using Deepgram nova-3 here too (germany) and noticed this too... Would be great so we could use it in other languages

GiteaMirror commented

2026-05-05 17:49:11 -05:00

@morbificagent commented on GitHub (May 9, 2025):

I temporarily tinkered around in
/app/backend/open_webui/routers/audio.py
Unfortunately, that's all I can do for now, but it works at the moment until (hopefully) it gets properly implemented.

            #this replaces the original params-part
            params = {
                "smart_format": "true",
                "detect_language": "true",
                "model": "nova-3"
            }

            # Make request to Deepgram API
            r = requests.post(
                "https://api.deepgram.com/v1/listen",
                headers=headers,
                params=params,
                data=file_data,
            )
            r.raise_for_status()
            response_data = r.json()

@morbificagent commented on GitHub (May 9, 2025): I temporarily tinkered around in /app/backend/open_webui/routers/audio.py Unfortunately, that's all I can do for now, but it works at the moment until (hopefully) it gets properly implemented. ``` #this replaces the original params-part params = { "smart_format": "true", "detect_language": "true", "model": "nova-3" } # Make request to Deepgram API r = requests.post( "https://api.deepgram.com/v1/listen", headers=headers, params=params, data=file_data, ) r.raise_for_status() response_data = r.json() ```

GiteaMirror commented

2026-05-05 17:49:14 -05:00

@blazejp83 commented on GitHub (May 14, 2025):

+1, interested in this as well

@blazejp83 commented on GitHub (May 14, 2025): +1, interested in this as well

GiteaMirror commented

2026-05-05 17:49:18 -05:00

@tjbck commented on GitHub (May 23, 2025):

#13989

@tjbck commented on GitHub (May 23, 2025): #13989

GiteaMirror commented

2026-05-05 17:49:22 -05:00

@jfouret commented on GitHub (Jun 19, 2025):

@tjbck I think it works only with Whisper

@jfouret commented on GitHub (Jun 19, 2025): @tjbck I think it works only with Whisper

GiteaMirror commented

2026-05-05 17:49:24 -05:00

@homjay commented on GitHub (Jun 28, 2025):

I'm still having a little trouble choosing a language while using the Deepgram Voice API. Am I missing something?

@homjay commented on GitHub (Jun 28, 2025): I'm still having a little trouble choosing a language while using the Deepgram Voice API. Am I missing something?

GiteaMirror commented

2026-05-05 17:49:27 -05:00

@homjay commented on GitHub (Jun 28, 2025):

@tjbck I think it works only with Whisper

Perhaps, we should create another issue

@homjay commented on GitHub (Jun 28, 2025): > [@tjbck](https://github.com/tjbck) I think it works only with Whisper Perhaps, we should create another issue

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/open-webui#55673