Upload a PDF, Word doc, or text file and see how many tokens it costs — and which models can actually hold it in context. Nothing is uploaded to a server; everything runs locally in your browser.
Your document
📄
Drop a file here, or click to browse
PDF, Word (.docx), or plain text — processed entirely in your browser
✕
Estimated tokens
0
Characters
0
Words
0
Pages (est.)
—
How this is estimated: ~4 characters per token, the standard quick-sizing rule of thumb. Exact tokenization varies by model.
Model compatibility
"Tight" means the document alone uses most of the context window, leaving little room for the model's response or conversation history.