Settings & Preferences

Open settings from the workspace selector dropdown in the sidebar — click your workspace name, then select Settings.

Profile

The Profile tab lets you customize your identity:

• Display Name — How your name appears to teammates in shared conversations
• Email — Your login email (read-only)
• Role — Your role in the current workspace (read-only)
• Timezone — Used for scheduling and time-related responses
• Theme — Choose between System (follows your OS), Light, or Dark

Click Save after making changes.

Model selection

The Model tab lets you choose which AI models the agent uses:

• Default model — The primary model for chat conversations
• Fast model — A faster model for quick responses when selected
• Reasoning model — A model optimized for complex, multi-step reasoning

Your admin may have pre-configured these, but you can adjust them to your preference. The dropdown shows available models from configured providers.

You can also adjust:

• Max iterations — How many tool-call loops the agent can run per response
• Max input/output tokens — Limits on conversation and response length

Tip

If you’re unsure which model to pick, the defaults work well for most tasks. You might switch to the fast model for simple questions and the reasoning model for complex analysis.

Usage analytics

The Usage tab shows how your workspace is using the platform:

• Cost over time — A chart showing daily, weekly, or monthly spending
• Token breakdown — Input tokens, output tokens, cache reads, and cache creation
• Model usage — Which models are being used and how much each costs
• Activity stats — Total LLM calls, average response time, and conversation count

Use the period selector (Day / Week / Month / All) to adjust the time range.

MCP connection

The Profile tab includes a MCP Connection card with your workspace ID. If you want to connect an external client (like Claude Desktop or VS Code) to your NimbleBrain instance, you’ll need this ID. See Connecting External Clients for setup instructions.