openaienmodel: gpt-5-mini-2025-08-07
Working with files in ChatGPT
Key Points
- Upload and edit many file formats
- Generate tables, charts, and downloadable outputs
- Connect third‑party apps with admin controls
Summary
ChatGPT lets you upload files directly into a chat to analyze, edit, summarize, extract data, and generate new outputs without leaving the conversation. Supported formats include CSV, XLSX, PDF, DOCX, JPEG, PNG, TXT and more. You can request views (tables, charts), edit documents, extract structured data, and download ChatGPT‑generated outputs.
Key Points
- How to upload: open a chat → Tools → “Add photos or files”.
- Supported formats: CSV, XLSX, PDF, DOCX, TXT, JPEG, PNG, etc.
- Common tasks: summarize reports, extract dates/owners, clean or rewrite documents, visualize sales data, convert PDFs to tables or spreadsheets.
- Output options: request tables, charts, or updated files and download them (updated spreadsheets, PDFs, etc.).
- Apps integration: enable via Settings → Apps, complete auth/permissions, apps appear in Tools or via @ or /. Use apps to bring external context into a chat.
- Enterprise notes: workspace admins control available apps; data accessed through apps is not used to train OpenAI models by default. Check File Uploads FAQ and Retention Policies for org-specific rules.
Practical tips for engineers
- Be explicit in prompts: specify output format (CSV, table, chart) and any schema or columns you need.
- For large datasets, ask for summaries or sample queries first before full transformations.
- Verify sensitive data handling with your Org admin and consult retention policies before sharing proprietary files.