[GH-ISSUE #3579] [FEAT]: support for wizper v3 from fal.ai for Speech recognition #2311

Open
opened 2026-02-22 18:29:07 -05:00 by yindo · 2 comments
Owner

Originally created by @MyButtermilk on GitHub (Apr 1, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3579

What would you like to see?

The currently supported STT models are slow and inaccurate. According to artificial analysis wiper v3 from fal.ai would be extremely fast, cheap and still accurate.

You could also support Eleven Labs STT via fal.ai. It is more expensive but also more accurate.

Originally created by @MyButtermilk on GitHub (Apr 1, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3579 ### What would you like to see? The currently supported STT models are slow and inaccurate. According to artificial analysis wiper v3 from fal.ai would be extremely fast, cheap and still accurate. You could also support Eleven Labs STT via fal.ai. It is more expensive but also more accurate.
yindo added the enhancementIntegration Requestfeature request labels 2026-02-22 18:29:07 -05:00
Author
Owner

@MyButtermilk commented on GitHub (Apr 12, 2025):

Any update on this feature request?

@MyButtermilk commented on GitHub (Apr 12, 2025): Any update on this feature request?
Author
Owner

@timothycarambat commented on GitHub (Apr 14, 2025):

We prioritize general GitHub integrations based on "upvotes" or activity/demand on a GH issue. If you need something delivered on your timeline, it is better suited via a self-maintained fork.

This issue has no activity aside from the main author, so it would fall under low priority. As you can imagine, we need to balance historical code, features, and integrations carefully and cannot just start working on every integration request we get instantly if the overall impact is minimal to the whole project.

If overwhelming, there is demand for a feature or integration, like how MCP was upvoted a ton on its issue, we will then make a point to add it into the core product. Otherwise, we have many people who fork and do their more custom or bespoke work so they can extend the core product into their unique needs.

@timothycarambat commented on GitHub (Apr 14, 2025): We prioritize general GitHub integrations based on "upvotes" or activity/demand on a GH issue. If you need something delivered on your timeline, it is better suited via a self-maintained fork. This issue has no activity aside from the main author, so it would fall under low priority. As you can imagine, we need to balance historical code, features, and integrations carefully and cannot just start working on every integration request we get instantly if the overall impact is minimal to the whole project. If overwhelming, there is demand for a feature or integration, like how MCP was upvoted a ton on its issue, we will then make a point to add it into the core product. Otherwise, we have many people who fork and do their more custom or bespoke work so they can extend the core product into their unique needs.
yindo changed title from [FEAT]: support for wizper v3 from fal.ai for Speech recognition to [GH-ISSUE #3579] [FEAT]: support for wizper v3 from fal.ai for Speech recognition 2026-06-05 14:45:49 -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#2311