Google's AI chatbot Gemini has received a series of recent updates, including new animations and other visual tweaks, alongside performance improvements in Gemini 3.5. At its recent I/O developer conference the company introduced the concept of 'Thinking' levels for the chatbot, offering two options: Normal and Expanded.
What Normal and Expanded mean
The Normal setting is optimized for faster, more concise replies. The Expanded setting allows Gemini to dig deeper into a topic and present more detailed, chain-of-thought style reasoning before replying. This deeper processing takes additional time to generate responses.
Wider availability and how to access it
After an initial, limited rollout, the Thinking levels are now being made broadly available; this wider deployment has been reported externally and we were able to confirm it ourselves. On the web interface, users can click the “flash” label at the right of the text box and choose between the two options under the “thinking level” menu.
The feature is free for all users and does not require a subscription. It is available not only on the web but also within Gemini's mobile apps.
Limits and practical considerations
Users should be aware that using the Expanded mode can cause them to reach usage limits sooner; when limits are exceeded, Gemini may temporarily restrict access. Choosing Expanded therefore trades faster responses for greater depth and a higher rate of quota consumption.
In short: Gemini's Thinking levels are now widely accessible and free on both web and mobile, offering a choice between speed and depth—while the deeper mode may lead to slower replies and quicker consumption of usage allowance.
