Licensed to be used in conjunction with basebox, only.
// ai chat
Maximum text length
Processing content that's too large
If an error message appears (e.g., "Document too large"), this is usually due to the context window.
What is the context window?
In basebox, the amount of content that can be processed is limited by what's known as the context window. This technical limitation of the AI determines how much text can be analyzed at once.
Important: This limit differs from file size (e.g., MB) and reflects a current constraint of the system.
The problem: "Document too large" error
When this error appears, it's usually because the context window limit has been reached, not because of the file size.
How to solve the problem:
Immediate solutions:
- Split the document – Break the document into smaller sections
- Select relevant content – Upload only the most important parts
- Copy text – Copy specific text instead of uploading entire files
Understanding context window vs. file size
The key difference:
The limit is based on the amount of content (tokens), not the file size.
In practice, this means:
- A small PDF (2 MB) can be too large if it contains a lot of text
- A large Excel file (10 MB) may work fine if it contains little text
Special file types:
Excel files:
- Every individual cell counts as content
- Even empty cells can count against the limit
PDF files:
- Can contain a lot of text even if they appear small
- Especially problematic for text-heavy documents
Rough guidelines for file sizes
Typical reference values (may vary):
- PDF: approx. 50 pages
- Excel: approx. 500 rows
- Text: approx. 100,000 characters
Important
These are only guidelines and can vary significantly depending on structure and content.
Optimizing content for basebox
Before uploading:
Optimizing Excel files:
- Remove empty rows and columns
- Delete unneeded worksheets
- Focus on the relevant data
Optimizing PDF files:
- Extract only the relevant pages
- Use targeted copy-paste instead of a full upload
General tips:
- Work with several smaller uploads instead of one large file
- Use only the content you actually need to analyze
Workarounds for large documents
If you need to process large documents:
- Section-by-section analysis – Split the document and analyze each part separately
- Targeted extraction – Copy only the relevant text passages
- Multiple chats – Use different chats for different parts of the document
- Create a summary – Have basebox summarize sections
Important note
The "Document too large" error usually refers to the context window, not the file size. Reducing the file size alone often does not solve the problem.
Still having trouble with large documents? Contact support