Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 22:23:50 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #2121] [BUG]: Scraping with agent does not leverage full context window #1381

New Issue
Closed
opened 2026-02-22 18:24:32 -05:00 by yindo · 0 comments
yindo commented 2026-02-22 18:24:32 -05:00
Owner
Copy Link
Copy Source

Originally created by @mikaeluu85 on GitHub (Aug 15, 2024).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2121

Originally assigned to: @timothycarambat on GitHub.

How are you running AnythingLLM?

AnythingLLM desktop app

What happened?

An error is presented:

"This page's content is way too long. I will summarize it right now."

Expected outcome is full scrape, the page is limited in length and should fit in contect window (128k)

Per discussions on Discord with @timothycarambat:

Tim Carambat — Yesterday at 9:43 PM
Your model's context window is too small for the content - so it overflows and needs to be "summarized" so the model can digest it
What model, provider?
Mikael — Yesterday at 9:53 PM
Ah, so its model-related. I thought is was anythingllm related (desktop).
OpenAI gpt4o in this case.
But, isnt gpt4o like 128k?
Do the scrape use like html2text or similar lib?
Or is the contect all html?
Mikael — Yesterday at 10:00 PM
context*
Tim Carambat — Yesterday at 10:14 PM
its like html2text, but if you are using gpt-4o that is plenty of space for sure
Ah i see where this is going on:
https://github.com/Mintplex-Labs/anything-llm/blob/bdf37bf180a7920bff93c5dd75597d9a7ad2519a/server/utils/agents/aibitat/providers/ai-provider.js#L176
Mikael — Yesterday at 10:25 PM
Aha, the 8k I believe is the limit for non paid, no?
Or atleast I have read that somewhere.
Tim Carambat — Yesterday at 10:26 PM
Not anymore, it used to be the OpenAI limit but that does not apply anymore

Are there known steps to reproduce?

Using AnythingLLM desktop client, Version 1.6.2 (1.6.2) and macOS Sonoma (14.6.1).
Provider used: OpenAI with Paid API
LLM: GPT-4o and 3.5-16k

Invoke the agent scraper on for example:

https://www.deskbird.com/blog/office-space-planning

Originally created by @mikaeluu85 on GitHub (Aug 15, 2024). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/2121 Originally assigned to: @timothycarambat on GitHub. ### How are you running AnythingLLM? AnythingLLM desktop app ### What happened? An error is presented: "This page's content is way too long. I will summarize it right now." Expected outcome is full scrape, the page is limited in length and should fit in contect window (128k) Per discussions on Discord with @timothycarambat: **Tim Carambat — Yesterday at 9:43 PM** Your model's context window is too small for the content - so it overflows and needs to be "summarized" so the model can digest it What model, provider? **Mikael — Yesterday at 9:53 PM** Ah, so its model-related. I thought is was anythingllm related (desktop). OpenAI gpt4o in this case. But, isnt gpt4o like 128k? Do the scrape use like html2text or similar lib? Or is the contect all html? **Mikael — Yesterday at 10:00 PM** context* **Tim Carambat — Yesterday at 10:14 PM** its like html2text, but if you are using gpt-4o that is plenty of space for sure Ah i see where this is going on: [https://github.com/Mintplex-Labs/anything-llm/blob/bdf37bf180a7920bff93c5dd75597d9a7ad2519a/server/utils/agents/aibitat/providers/ai-provider.js#L176](https://github.com/Mintplex-Labs/anything-llm/blob/bdf37bf180a7920bff93c5dd75597d9a7ad2519a/server/utils/agents/aibitat/providers/ai-provider.js#L176) **Mikael — Yesterday at 10:25 PM** Aha, the 8k I believe is the limit for non paid, no? Or atleast I have read that somewhere. **Tim Carambat — Yesterday at 10:26 PM** Not anymore, it used to be the OpenAI limit but that does not apply anymore ### Are there known steps to reproduce? Using AnythingLLM desktop client, Version 1.6.2 (1.6.2) and macOS Sonoma (14.6.1). Provider used: OpenAI with Paid API LLM: GPT-4o and 3.5-16k Invoke the agent scraper on for example: https://www.deskbird.com/blog/office-space-planning
yindo added the enhancement label 2026-02-22 18:24:32 -05:00
yindo closed this issue 2026-02-22 18:24:32 -05:00
yindo changed title from [BUG]: Scraping with agent does not leverage full context window to [GH-ISSUE #2121] [BUG]: Scraping with agent does not leverage full context window 2026-06-05 14:40:28 -04:00
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
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#1381
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?