[GH-ISSUE #4440] [DOCS]: vector database and embeder token vs characters #2832

Closed
opened 2026-02-22 18:31:27 -05:00 by yindo · 2 comments
Owner

Originally created by @kalle07 on GitHub (Sep 26, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4440

Description

your menue is not consistant!
you show up "token" (more or less) in LLM and in embedder!
but if i choose max text-chunk there is written characters ? what is it?

Image

because the max number i can set here is max token from embedder settings, so it is not consitant!

Originally created by @kalle07 on GitHub (Sep 26, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4440 ### Description your menue is not consistant! you show up "token" (more or less) in LLM and in embedder! but if i choose max text-chunk there is written characters ? what is it? <img width="470" height="408" alt="Image" src="https://github.com/user-attachments/assets/0c443480-ee99-44f9-a89c-cc343b44a996" /> because the max number i can set here is max token from embedder settings, so it is not consitant!
yindo added the documentation label 2026-02-22 18:31:27 -05:00
yindo closed this issue 2026-02-22 18:31:27 -05:00
Author
Owner

@timothycarambat commented on GitHub (Sep 29, 2025):

It is character because, unless we run the models' specific tokenizer, we cannot produce an accurate count - the hint in both does not mention token to avoid that confusion

@timothycarambat commented on GitHub (Sep 29, 2025): It is character because, unless we run the models' specific tokenizer, we cannot produce an accurate count - the hint in both does not mention token to avoid that confusion
Author
Owner

@kalle07 commented on GitHub (Sep 30, 2025):

so my question is ... thats also chararcters because its the max number i can set in text-chunk

Image

here is not mentioned that this are characters.

For the complicated calculation, you could perhaps at least provide an estimated number of token, since characters are independent of language words.

For example, 5000-6000 characters correspond to approximately one page and approximately 1200-1400 token.

@kalle07 commented on GitHub (Sep 30, 2025): so my question is ... thats also chararcters because its the max number i can set in text-chunk <img width="929" height="517" alt="Image" src="https://github.com/user-attachments/assets/4177d7c6-2321-42e8-93e6-24e9eaa6f18e" /> here is not mentioned that this are characters. For the complicated calculation, you could perhaps at least provide an estimated number of token, since characters are independent of language words. For example, 5000-6000 characters correspond to approximately one page and approximately 1200-1400 token.
yindo changed title from [DOCS]: vector database and embeder token vs characters to [GH-ISSUE #4440] [DOCS]: vector database and embeder token vs characters 2026-06-05 14:48:47 -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#2832