Ultrathink Has Its Own Spinner Now
When the model needs a moment to think harder, the terminal knows it.
Ultrathink has its own spinner. The feedback matching the operation — small detail that shouldn't matter, does anyway. AI CLI UX continues being refined.
Ultrathink — Claude Code's extended thinking mode, the one you invoke when you want the model to actually sit with a problem instead of firing back the first plausible answer — now has its own custom active render in the CLI.
Not the generic spinner. Its own thing.
This is a small detail that shouldn't matter, and it does anyway. When you're burning tokens on a hard problem and the terminal just shows the same animation it shows when it's fetching a diff, something feels off — like watching a surgeon and a vending machine make the same face. The feedback doesn't match the operation.
Now it does. The CLI signals that something different is happening. Which is either good UX design or anthropomorphization that costs money per second, depending on your mood.
Either way, the visual is there. The model is thinking. You can tell.
You still don't know what it's thinking about.
Counterpoints
Push back, extend the argument, or sharpen it. New counterpoints go through review before they show up here.
No approved counterpoints yet.