Documents, Images, Knowledge Collections, and Integrations

You can share documents, images, and knowledge collections in chat so the AI can analyze, summarize, or use them as a basis for its responses. You can also connect integrations to use external services directly in the chat.

Adding Documents

Supported Formats

Format	File Type	Typical Use
PDF	.pdf	Reports, decisions, forms
Word	.docx	Letters, templates, drafts
PowerPoint	.pptx	Presentations
Excel/CSV	.xlsx, .xls, .csv	Tables, lists, data
Audio	.mp3, .m4a, .wav, .webm	Meeting recordings, voice memos, interviews

Audio Files

When you upload an audio file, it is automatically transcribed so the AI can work with the spoken content as text. This means you can ask questions about meetings, interviews, or voice memos just like you would with a written document.

Limits

Limit	Value
Maximum number of documents per chat	20
Maximum file size	25 MB

If you reach the document limit, remove a document before adding another. For larger collections of files, consider using a knowledge collection instead.

How to Add a Document

1. Click the plus icon (+) in the input area
2. Select Upload document
3. Select one or more files from your computer — you can select multiple files at once
4. Wait until the files are uploaded and processed — the send button is disabled until processing is complete
5. Ask your question about the documents

Processing

After you upload a document, it is automatically processed so the AI can read it. You cannot send your message until processing is complete. This usually takes only a few seconds.

Practical Example: Summarize Document

You have a 30-page PDF of a committee meeting minutes. Upload it to the chat and write:

“Summarize these minutes. List the most important decisions and open tasks with responsible persons.”

Adding Images

Supported Formats

JPEG, PNG, GIF, and WebP – maximum 10 MB per image, up to 8 images per message.

How to Add Images

1. Click the plus icon (+) in the input area
2. Select Upload image
3. Select the desired image from your computer

What the AI Can Do with Images

• Describe content – e.g., explain what’s in a photo
• Read text in images – e.g., extract text from a photo or screenshot
• Analyze – e.g., evaluate diagrams or plans

Practical Example: Analyze Image

You receive a photo of a damaged playground attached to a defect report. Upload the image to the chat and ask:

“Describe the visible damage in this photo. Create a list of defects for the public works department.”

Generating Images with AI

You can also have the AI create your own images — see Generating Images with AI.

Adding a Knowledge Collection

You can add a knowledge collection directly to a new chat so the AI can access the documents it contains:

1. Start a new chat
2. Click the plus icon (+) in the input area
3. Select the desired knowledge collection

The AI can then answer questions based on the documents in that knowledge collection — without you having to upload the files individually.

Tips for Working with Files

• Attach relevant files – Only upload documents relevant to your question
• Ask targeted questions – “Summarize pages 5–8” is better than “What’s in the document?”
• Consider data protection – Check if the document contains personal data and use anonymous mode if needed

Adding an Integration

You can connect integrations (external services and tools) directly to an active chat, even without using a skill:

1. Click the + button next to the input field
2. Select Add Integration
3. Choose the desired integration from the submenu
4. The integration is now available for the AI in this chat

The AI can then access the connected service — for example, to retrieve data from a specialized system or perform actions on your behalf.

Tip

You can add multiple integrations to the same chat. Already attached integrations are automatically hidden from the list. If an integration shows as unavailable, you may need to authorize it first under Settings → Integrations.

Next Steps

• Knowledge Collections – Create and manage knowledge collections
• Anonymous mode – Protect personal data in documents