Skip to Content

AI Chat

AIDEN uses the same core chat experience across the product. What changes is not the chat system — it is the context:

  • project chat carries project-level context, for broader planning
  • story chat stays anchored to one story and its spec, for focused execution
  • agent chat reflects the role and instructions of that agent

Use the chat closest to the work you want to move. The closer the context matches the task, the better the output.


What you can do in chat

  • ask questions about the current work
  • plan a task or refine a spec
  • review implementation choices
  • attach files to give the model exactly the material it needs
  • pick the model for the conversation when a task calls for a different tradeoff — see AI Providers

Multi-pane chat

For heavier sessions, AIDEN has a fullscreen chat workspace:

  • run 1 to 4 chat panes side by side
  • manage sessions with tabs, so parallel threads stay organized
  • spawn agents directly from a pane when work should be delegated
  • choose models and attach files per pane

This is the surface for the moments when one conversation is not enough — comparing two approaches in parallel, or keeping a planning thread open while an implementation thread runs.


Prefer talking?

Everything chat can start, voice can start too. Talk to AIDEN is a push-to-talk voice layer over the same system — quick actions run instantly, heavier requests are delegated to a background agent.


The practical mental model

Do not think of AIDEN as having separate chat products. It is one AI chat layer that becomes more useful when attached to the right context. That is what keeps conversations close to real software work instead of turning everything into one giant generic thread.