NUFI Docs

Files and images

Upload documents, send images, and ask the AI questions about them.

You can attach files to a message and ask the AI about them — read a PDF, describe a photo, turn a screenshot into a table.

Attach a file

Click the paperclip (Attach File) at the bottom-left of the message box. You can also just drag a file onto the conversation.
Choose how to send it: Upload to Provider (sends the file to the model — best for images and rich documents) or Upload as Text (pulls the text out and adds it to your prompt).
Type your question and press Enter. The file rides along with the message.

The attach menu with Upload to Provider and Upload as Text

You can attach more than one file to a single message.

Search across your files

Open the Tools menu (the sliders icon next to the paperclip) and turn on File Search. The AI can then look through the files you've uploaded and pull in the relevant parts when it answers — you don't have to paste the contents yourself.

The Tools menu — File Search, Web Search, Skills, Run Code, Artifacts

Talk to a document

Upload a PDF, then ask questions about it:

Summarise the main findings of this report in 5 bullets.

What does section 3.2 say about pricing?

Translate the conclusion into Vietnamese.

NUFI feeds the relevant parts to the model and answers in context.

Look at an image

Upload an image, then ask:

What is in this photo?

Read the text in this screenshot and turn it into a table.

Suggest three improvements to this UI design.

This needs a vision-capable model. If the model can't read the image, switch models from the model picker.

What works

TypeExamples
ImagesPNG, JPEG, WebP, GIF
DocsPDF, plain text, Markdown
Code.py, .ts, .go, .md, …

Typical limits per message (your admin can change these):

LimitDefault
Files / msg5
Per file size20 MB
Total50 MB

Generate and edit images

If image generation is enabled for your team, you can create images from a description. Turn on the image tool from the Tools menu (or type /image at the start of the box if your deployment uses that), then describe what you want:

a watercolour fox reading a book in a cosy library,
soft pastel colours, dawn lighting

The image appears in the conversation and is saved to your account. To change it, describe the edit in your next message ("make the sky purple and add a moon") — you keep both versions.

If you don't see an image tool, your organisation hasn't turned on image generation. Ask your admin.

Storage and privacy

Files you upload are saved to your account and visible only to you (plus anyone you explicitly share a conversation with). Your own files are listed under the avatar menu → My Files. Deleting a conversation also deletes the files attached to it.

Errors you might see

  • "File too large" — over the per-file cap. Compress or split it.
  • "Out of budget" — image generation costs more than text. Check your usage in the console.
  • "Please slow down" — too many requests in a short window. Wait a few seconds.