📁 File Block

Overview

The File Block allows you to use the complete content of one or more documents as context for AI.

Unlike the Search Block which sections documents to exploit specific extracts, the File Block provides the entire content to the AI, which is particularly useful for short documents or uses requiring global context understanding.

Operation and specifications

Using files

The File Block offers two options for adding documents, which can be used separately or combined:

File attribute

Similar to the Search Block, you can use a File attribute that will be populated by the User when creating a publication. This file attribute will be populated by a Wizard step.

Static files

You can directly add "static" documents via the block parameters. These files will always be available for the block, without requiring user action.

File attribute configuration

In the File Block parameters, you must specify the File Attribute that will be used.

This File Attribute must be associated with a File Upload Step in the Wizard so that the User can upload the file(s) that will populate this File Attribute.

Create File Attribute

Define a File type attribute in your App's attribute menu

Configure Block Parameters

In the File Block settings, select the File Attribute you created

Set Up Wizard Step

Add a File Upload step in your Wizard and link it to your File Attribute

Supported file types and limits

Supported formats:

PDF (.pdf)

Word (.docx)

PowerPoint (.ppt)

Excel (.xlsx)

CSV (.csv)

File limitations:

Maximum file size: 50 MB per file

Context Limitations

⚠️ The main limitation comes from the maximum context accepted by AI models. It is recommended to use relatively short documents and/or AI models that accept larger contexts so that the prompt is accepted. You can consult the documentation on 🤖 AI Block to learn more about models suitable for larger contexts.

Integration with AI blocks

The File Block integrates into AI blocks in the same way as the Search Block. The document content will be used as context for the AI during content generation.

📚 Related documentation: The documentation on 🤖 AI Block.

Use cases and best practices

The File block is particularly useful in the following situations:

📖 Complete analysis

Analysis of short documents requiring global understanding

🎯 Full context

Processing documents where complete context is essential

🔍 Beyond extracts

Situations where using extracts (like with the Search Block) doesn't allow correct interpretation of information

🏗️ Complex structure

Documents with complex structures that need to be understood as a whole

Best practices:

Use the File block for documents requiring complete analysis

Combine file attributes and static documents for more flexibility

Pay attention to AI model context limits when using large documents

Choose appropriate AI models based on your document size and complexity needs

Workflow optimization:

Document assessment

Evaluate if your document needs complete analysis or if extracts would suffice

Model selection

Choose an AI model on the AI block with sufficient context capacity for your document size

Configuration setup

Configure both static files and dynamic file attributes as needed

Testing & refinement

Test with various document sizes and refine your approach based on results

File block vs Search block

Feature	File Block	Search Block
Content usage	Complete document	Specific extracts
Best for	Short docs, full context needed	Large docs, targeted information
Context impact	Higher context consumption	Lower context consumption
Processing speed	Slower with large files	Faster, more efficient

Performance considerations

Document size: Larger documents consume more context tokens

Model choice: Use high-context models (like Gemini Pro or GPT4.1) for large documents

Processing time: Complete document analysis may take longer than extract-based processing

Context efficiency: Consider if full document analysis is necessary for your use case

Overview#

Operation and specifications#

Using files#

File attribute#

Static files#

File attribute configuration#

Supported file types and limits#

Supported formats:#

File limitations:#

Integration with AI blocks#

Use cases and best practices#

The File block is particularly useful in the following situations:#

Best practices:#

Workflow optimization:#

File block vs Search block#

Performance considerations#