Skip to content

Chat Interface

The chat interface is the main way to interact with AI models in Catalyst. It provides a familiar conversational experience with powerful features for customization and control.

  1. Navigate to Chat from the sidebar menu
  2. Click the + button to create a new chat session
  3. Type your message in the input box and press Enter or click Send

Use the model dropdown in the chat input area to switch between available AI models. Each model has different capabilities:

  • VooV - Default model with GPT-4o-mini capabilities
  • Custom models - Models you’ve imported from OpenAI, Ollama, or other providers

To add more models, see the Importing Models guide.

Click the Settings icon in the chat header to access:

Controls how creative or deterministic responses are:

  • Lower values (0.0-0.3) - More focused, consistent responses
  • Higher values (0.7-1.0) - More creative, varied responses

Set a custom system prompt to give the AI specific instructions or a persona. This message is sent at the start of every conversation.

Responses stream in real-time as they’re generated. You can see the AI “thinking” as it writes.

Code in responses is automatically syntax highlighted. Click the copy button to copy code to your clipboard.

The AI can generate diagrams using Mermaid syntax. These render automatically as visual diagrams.

LaTeX math expressions are rendered properly using KaTeX.

With vision-capable models, you can attach images to your messages for the AI to analyze.

The sidebar shows all your chat sessions. Click on any session to continue that conversation.

Use the search box to find past conversations by content.

Create a new session from any message in your history. This lets you explore different directions from a specific point in a conversation.

Share a read-only view of your chat session with others:

  1. Click the Share button in the chat header
  2. Copy the generated link
  3. Anyone with the link can view the conversation
  • Create new sessions when switching topics to keep context clean
  • Use lower temperature for factual questions, higher for creative tasks
  • Long conversations use more context - start fresh if responses become inconsistent