Skip to content
Service status

Type to search

Functionality · 1 min read

Documents and artifacts

Documents and visualisations the assistant creates — how they appear, how to iterate, and how to get them out.

An assistant can answer with text in the chat, but it can also create documents — called artifacts in Intric. Every artifact you or your assistants have created is collected under Artifacts in the left menu.

Document creation is not a feature you switch on. It is a capability of the language models, triggered either when the model judges a document to be the appropriate response format, or when you explicitly ask for one (“write this up as a document”). Different models use the capability to different degrees — if you did not get a document when you wanted one, ask for it outright.

Once an artifact exists, you can work on it two ways:

  • Ask the assistant to change it — “shorten section two”, “add a summary at the top”.
  • Edit it yourself by clicking the pencil.

A finished document can be saved as PDF or DOCX.

You can also ask an assistant to build a visualisation — including an interactive one, with clickable elements. As with documents, this is triggered either by the model’s own judgement or by you asking (“visualise this”, “turn it into a presentation”). Visualisations download as source code, which you can then open and use.

When you run a Plan, its steps can create and update artifacts during the run: step 1 creates a document, step 2 picks it up and builds on it. Note that artifacts created by a Plan run cannot currently be edited directly or carried into an ordinary conversation — to change the result, adjust the Plan instructions and run it again.

Read more in the help center: Functionality and use cases.

Test your knowledge

Question 1 of 2

Question 1 of 2

What is an artifact in Intric?

What is an artifact in Intric?