File and document analysis

Delos allows you to analyze your files to extract information, summarize them, answer your questions, or automate tasks. The AI adapts to many formats and professional use cases.

Supported formats

Notes:

• PDFs must contain selectable text (not image scans without OCR).
• Image analysis (description, text extraction, etc.) requires a multimodal model (e.g., GPT-4o, Gemini).
• Code files are read, analyzed, and commented on, but not executed or compiled.

How to analyze a file with Delos

1. Click the attachment icon in the message area.
2. Select your file from your computer.
3. Ask your question in the text area:
   • "Summarize this document"
   • "Extract key figures"
   • "Describe this image"
   • "Analyze this code"
4. Send: the AI analyzes the content and responds directly.

Common use cases

• Document summary: Get a clear synthesis of a report, contract, or large document.
• Data extraction: Retrieve figures, tables, or KPIs from an Excel or CSV file.
• Image analysis: Describe, analyze, or extract text from an image (with multimodal model enabled).
• Code review: Receive improvement suggestions, comments, or a summary of your code.
• Document translation: Translate a complete document in Office format (DOCX, PPTX, ODT, etc.).

⚠️ Scanned PDFs are not translated (no native OCR support).

Limits and precautions

• Maximum file size: 50 MB
• Multimodal models required: Image analysis only possible with a multimodal model (GPT-4o, Gemini, etc.).
• Credit consumption: Analyzing large or complex files may consume more credits depending on the model used.
• OCR not supported: Scanned PDFs or images containing non-selectable text are not analyzed or translated.

Tip: For best results with PDFs, ensure they contain selectable text (avoid image scans). For image analysis, verify that the multimodal model is enabled in your Delos environment.