[GH-ISSUE #4070] [FEAT]: Add TTS + STT from AnythingLLM to the embed widget #2593

Open
opened 2026-02-22 18:30:22 -05:00 by yindo · 2 comments
Owner

Originally created by @Alminc91 on GitHub (Jun 30, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4070

What would you like to see?

Hello community.

What I really love in the existing AnythingLLM app is the ability to add TTS and STT.

Would it be possible to reuse this configuration and add it to the chatbot embed? So that you could also talk to the chatbot and get the response via speech output, especially for disabled people or something.

Would that be very difficult and a lot of work to integrate? Otherwise I would be interested in adding that as well. But if there are some starting hints and tips on how I could do that, that would be awesome.

Originally created by @Alminc91 on GitHub (Jun 30, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4070 ### What would you like to see? Hello community. What I really love in the existing AnythingLLM app is the ability to add TTS and STT. Would it be possible to reuse this configuration and add it to the chatbot embed? So that you could also talk to the chatbot and get the response via speech output, especially for disabled people or something. Would that be very difficult and a lot of work to integrate? Otherwise I would be interested in adding that as well. But if there are some starting hints and tips on how I could do that, that would be awesome.
yindo added the enhancementfeature request labels 2026-02-22 18:30:22 -05:00
Author
Owner

@timothycarambat commented on GitHub (Jun 30, 2025):

Both of these would cause the supporting embed to be a much bigger bundle. TTS and STT are both done via WebWorkers and download suitable models for both functions and run client-side. The good part is that this doesnt not cause load on the server, the downside if the client has to absorb that totally and for a chat widget is likely too much bloat for such a simple interface tool

@timothycarambat commented on GitHub (Jun 30, 2025): Both of these would cause the supporting embed to be a **much** bigger bundle. TTS and STT are both done via WebWorkers and download suitable models for both functions and run client-side. The good part is that this doesnt not cause load on the server, the downside if the client has to absorb that totally and for a chat widget is likely too much bloat for such a simple interface tool
Author
Owner

@Alminc91 commented on GitHub (Jun 30, 2025):

@timothycarambat thanks for your explanation, I get it.

Maybe something like this would be feasible: chatbot icon is there with small bundle size. User clicks on icon, chat window opens -> WebWorkers and voice functions start to load in background at that point. And they use the settings from the main app.

So a delayed load that still keeps bundle small, when not needed or used. Like predictive loading. I will have to see if that would make sense, but would be a cool feature.

@Alminc91 commented on GitHub (Jun 30, 2025): @timothycarambat thanks for your explanation, I get it. Maybe something like this would be feasible: chatbot icon is there with small bundle size. User clicks on icon, chat window opens -> WebWorkers and voice functions start to load in background at that point. And they use the settings from the main app. So a delayed load that still keeps bundle small, when not needed or used. Like predictive loading. I will have to see if that would make sense, but would be a cool feature.
yindo changed title from [FEAT]: Add TTS + STT from AnythingLLM to the embed widget to [GH-ISSUE #4070] [FEAT]: Add TTS + STT from AnythingLLM to the embed widget 2026-06-05 14:47:27 -04:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2593