Test: Thinking Effort Level in Model Picker

[bhavyaus]

Refs: microsoft/vscode-copilot-chat#4515, #300235

Author: @sandy081 @bhavyaus

• Any OS @jrieken
• Any OS @mjbvz
• Any OS @eleanorjboyd

Complexity: 3

Create Issue

---

Thinking Effort Level in Model Picker

A new Thinking Effort submenu has been added to the model picker for models that support reasoning effort levels. The effort level controls how much reasoning/thinking the model applies to each request.

1. Picker Works:

   • Select a reasoning model (e.g., Claude Sonnet 4.6, GPT-5.4); verify a ">" arrow appears on the picker item
   • Click the arrow; verify a "Thinking Effort" submenu shows effort levels for that model
   • Select an effort level; verify the picker label updates (e.g., "Claude Sonnet 4.6 · High")
   • Select a non-reasoning model (e.g., GPT-4.1, GPT-4o); verify no submenu appears
   • Try a few different Anthropic and OpenAI models
   • Available levels vary per model (e.g., some have None, some skip Medium)

2. Effort Reaches the Backend: Select a reasoning model, pick an effort level, and send a message. Open the Chat Debug Logs and right-click on a log entry to view the raw chat request. Search for "effort" in the JSON; verify the value matches what you selected.

3. Persistence: Set an effort level on a model; start a new conversation with the same model; verify the effort level is retained.

4. Accessibility and Theming: Switch to High Contrast and Light themes; verify the effort submenu is readable and renders correctly. Use keyboard-only navigation to open the picker, select a model, and choose an effort level.

[sanket-bhalerao]

the Thinking effort selector does not appear on WSL #303982