Document
Drop a PDF or scanned image here
or use a sample document — everything runs in your browser, nothing is uploaded
Ask
Load a document (or pick a sample), then ask questions about it.
Answers are grounded in the document with clickable citations —
Auto uses OCR + layout + retrieval, Vision sends the
page image to a multimodal model.