AI tool
AI Token & Document Splitter
Estimate the size of a long document and divide it into copyable chunks, leaving room for the reply and overlapping context.
Character-based estimates do not guarantee that each chunk fits the target model.
Model context windowReserve room for the AI response before splitting.
Characters161characters
Words26words
UTF-8161bytes
Estimated tokens38approx.
Input limit used1%6,500 max
Document chunks1chunks
Document chunks (1)
Document budget
Leave response space and inspect each chunk
Subtract the response reserve
Set the total context window and the room you want for an answer. Leave extra space for your instructions and any existing conversation.
Adjust overlapping text
The splitter prefers paragraph, line, and sentence boundaries and repeats some text in the next chunk. More overlap increases the total material you send.
Review before transferring
Inspect the token estimate and ending of each chunk. Copy one, copy all, or download a TXT file.
A worked example
Reserve space in an 8K window
- Start with
- Context: 8,000 · response reserve: 1,500 · overlap: 200 tokens
- Check the result
- The input budget is 6,500 tokens. The text Hello is estimated at 2 tokens and remains one chunk.
The budget is for planning. Splitting uses an average character ratio, so check actual token counts separately.