Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-21 01:25:28 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #5224] [FEAT]: Transcription-only option in meeting assistant #4976

New Issue
Open
opened 2026-06-05 14:51:15 -04:00 by yindo · 5 comments
yindo commented 2026-06-05 14:51:15 -04:00
Owner
Copy Link
Copy Source

Originally created by @cwatl on GitHub (Mar 18, 2026).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/5224

What would you like to see?

The live transcription feature is working beautifully - and quickly. But then the transcript disappears while the local LLM takes a minute or two to generate a summary that I don't want. And only then am I able to see the transcription again.

Can we have an option to toggle the summary and q&a aspects off for if we only want the transcription and don't want to have to wait around for it?

Originally created by @cwatl on GitHub (Mar 18, 2026). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/5224 ### What would you like to see? The live transcription feature is working beautifully - and quickly. But then the transcript disappears while the local LLM takes a minute or two to generate a summary that I don't want. And only then am I able to see the transcription again. Can we have an option to toggle the summary and q&a aspects off for if we only want the transcription and don't want to have to wait around for it?
yindo added the enhancementfeature request labels 2026-06-05 14:51:15 -04:00
yindo commented 2026-06-05 14:51:15 -04:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Mar 18, 2026):

Hm, would exposing this as an api endpoint suffice? The intended purpose of the MA is for transcription/summary/tools

If we made a backend API for this, would that work for your use case? It seems you want the transcript for some specific output or use somewhere else?

<!-- gh-comment-id:4084492579 --> @timothycarambat commented on GitHub (Mar 18, 2026): Hm, would exposing this as an api endpoint suffice? The intended purpose of the MA is for transcription/summary/tools If we made a backend API for this, would that work for your use case? It seems you want the transcript for some specific output or use somewhere else?
yindo commented 2026-06-05 14:51:15 -04:00
Author
Owner
Copy Link
Copy Source

@cwatl commented on GitHub (Mar 18, 2026):

I'm not sure if an API would meet my needs. I need the transcript as step one in a process where step two is much more complex than just a summary. There are many use cases where a transcript is an important element and an general AI summary isn't needed. Having that compute-intensive and time-intensive aspect that runs even when unneeded is the challenge for my usecase.

<!-- gh-comment-id:4085025521 --> @cwatl commented on GitHub (Mar 18, 2026): I'm not sure if an API would meet my needs. I need the transcript as step one in a process where step two is much more complex than just a summary. There are many use cases where a transcript is an important element and an general AI summary isn't needed. Having that compute-intensive and time-intensive aspect that runs even when unneeded is the challenge for my usecase.
yindo commented 2026-06-05 14:51:16 -04:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Mar 18, 2026):

Are you intending to just basically put the transcript directly into the chat? If so, we are already planning to use the transcription engine for audio uploads in the main application.

It sounds like thought you just need a UI to drag and drop stuff in an get a simple transcription, no fuss or API and then you copy that from AnythingLLM to your next step?

I think we can make something here like a toggle to "automatically process meetings" that is default on, but turning it off just does the transcript?

<!-- gh-comment-id:4085124364 --> @timothycarambat commented on GitHub (Mar 18, 2026): Are you intending to just basically put the transcript directly into the chat? If so, we are already planning to use the transcription engine for audio uploads in the main application. It sounds like thought you just need a UI to drag and drop stuff in an get a simple transcription, no fuss or API and then you copy that from AnythingLLM to your next step? I think we can make something here like a toggle to "automatically process meetings" that is default on, but turning it off just does the transcript?
yindo commented 2026-06-05 14:51:16 -04:00
Author
Owner
Copy Link
Copy Source

@cwatl commented on GitHub (Mar 18, 2026):

I think we can make something here like a toggle to "automatically process meetings" that is default on, but turning it off just does the transcript?

Yes, that would meet my needs exactly - and it's what similar applications tend to do in my experience, whether locally hosted or cloud based.

It sounds like thought you just need a UI to drag and drop stuff in an get a simple transcription, no fuss or API and then you copy that from AnythingLLM to your next step?

Essentially yes - though I'll be recording from within AnythingLLM rather than dropping something in. And ideally there's an option for the audio not to be saved.

Are you intending to just basically put the transcript directly into the chat? If so, we are already planning to use the transcription engine for audio uploads in the main application.

For my workflow, I have a Python script that targets a batch of transcripts directly where the files are stored -- it grabs one, serves it to LMStudio, waits for it to process, captures the output, deletes the transcript, and then repeats until the batch is fully processed - and in the end writes all of the outputs to a text file.

But that's just automating what in the past I would do with copying and pasting the way you described. I'm in an industry where many people will have my use case (psychology), and I anticipate most of them will simply copy and paste. I'm trying to educate mental health professionals and testing psychologists on the benefits of local LLMs rather than the big cloud services - and step one is a local way to record/listen to and transcribe audio in near real time.

<!-- gh-comment-id:4085660610 --> @cwatl commented on GitHub (Mar 18, 2026): >I think we can make something here like a toggle to "automatically process meetings" that is default on, but turning it off just does the transcript? Yes, that would meet my needs exactly - and it's what similar applications tend to do in my experience, whether locally hosted or cloud based. >It sounds like thought you just need a UI to drag and drop stuff in an get a simple transcription, no fuss or API and then you copy that from AnythingLLM to your next step? Essentially yes - though I'll be recording from within AnythingLLM rather than dropping something in. And ideally there's an option for the audio not to be saved. >Are you intending to just basically put the transcript directly into the chat? If so, we are already planning to use the transcription engine for audio uploads in the main application. For my workflow, I have a Python script that targets a batch of transcripts directly where the files are stored -- it grabs one, serves it to LMStudio, waits for it to process, captures the output, deletes the transcript, and then repeats until the batch is fully processed - and in the end writes all of the outputs to a text file. But that's just automating what in the past I would do with copying and pasting the way you described. I'm in an industry where many people will have my use case (psychology), and I anticipate most of them will simply copy and paste. I'm trying to educate mental health professionals and testing psychologists on the benefits of local LLMs rather than the big cloud services - and step one is a local way to record/listen to and transcribe audio in near real time.
yindo commented 2026-06-05 14:51:16 -04:00
Author
Owner
Copy Link
Copy Source

@cwatl commented on GitHub (Mar 18, 2026):

I should add that things seem to bug out after about 30 minutes of recording/transcribing. It stops showing new transcript content in the live transcript process and then, once concluded, it just hangs showing processing indefinitely. I waited ~30 minutes before giving up (even when on CPU only my computer will usually summarize a 60-minute meeting much faster than that). Not sure if it's just the audio transcription process overrunning something or if it's the summary element. Usually my meetings are about an hour and amount to ~12k to 18k tokens.

<!-- gh-comment-id:4085717009 --> @cwatl commented on GitHub (Mar 18, 2026): I should add that things seem to bug out after about 30 minutes of recording/transcribing. It stops showing new transcript content in the live transcript process and then, once concluded, it just hangs showing processing indefinitely. I waited ~30 minutes before giving up (even when on CPU only my computer will usually summarize a 60-minute meeting much faster than that). Not sure if it's just the audio transcription process overrunning something or if it's the summary element. Usually my meetings are about an hour and amount to ~12k to 18k tokens.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label enhancement feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#4976
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?