Product guide
How to upload documents and files to Whizi
Upload PDFs, Word docs, spreadsheets, images, and code to Whizi and chat with them using Claude, GPT, or Gemini.
Supported file types
- PDFs (text and scanned via automatic OCR)
- Word documents (.docx)
- Spreadsheets (.xlsx, .csv)
- Images (JPG, PNG, WebP) with multimodal models
- Code and text files (.py, .js, .ts, .go, .md, .txt, ...)
Best practices
Name your files clearly; the model sees the filename.
For long PDFs, use Gemini 3 Pro for fewer chunking issues.
For spreadsheets, ask the model to output a clean table or JSON.
For images, explicitly say "look at the attached image" to make sure the multimodal path is used.
Workflow checklist
- Pick the file
- Attach it to the chat
- Pick the best model for the file type
- Ask one question at a time
Common questions
Are my uploads private?
Yes. Files are stored securely in your workspace and never used to train models.
What’s the maximum file size?
Depends on the plan and the model. Gemini 3 Pro supports the largest files.
Can I delete a file after chatting?
Yes. Files can be removed from the workspace at any time.