Chat

The chat panel on Cortex Canvas gives you two modes: standard chat and CASi. Both live in the same interface, but they work from different scopes of context.

Standard chat

Standard chat is the focused option. Use it when you want full model choice and a tighter conversation around the material in front of you.

It is a good fit for:

• summarising or comparing specific material
• testing an idea quickly with a chosen model
• reasoning about a limited set of nodes without pulling in the wider client workspace

Standard chat keeps the thread history on that canvas, but it does not carry the same client-wide memory that CASi does.

CASi chat

CASi works across the full active client workspace. It can use your brand context, client notes, prior articles, research nodes, content history, competitors, keywords, training material, and previous CASi conversations across canvases for that client.

Use CASi when the output needs to carry your brand voice, find something in the client account, report on what already exists, or draft from a broader body of context than the current canvas alone.

Features available in both modes

• Model selection. Switch between different AI models mid-conversation. The task gets matched to the right model without copy-paste.
• Web search. Capture live sources in-thread. Current claims stay testable.

Both chat modes are available to any user with positive AI balance.

Memory and scope

Standard chat remembers the current thread on that canvas.

CASi remembers previous CASi conversations across canvases inside the same client account. That memory is client-scoped, so one client's history never bleeds into another.

Using chat alongside the canvas

Chat and the canvas work together. You can ask standard chat to reason through a specific source cluster, ask CASi to draft from the wider workspace, or turn a chat result into a new node that feeds back into the canvas.