Overview
- Google began rolling out Thinking Levels across the web, Android and iOS on Wednesday, June 3, making Standard and Extended modes available to free and paid users in select markets including India.
- Extended Thinking makes Gemini pause and run longer, step‑by‑step reasoning before replying so users get more detailed explanations and stronger problem solving for complex requests.
- The new controls apply to specific models — Gemini 3.1 Flash‑Lite, Gemini 3.5 Flash and Gemini 3.1 Pro — and some model variants support up to three thinking levels.
- A highest tier called Deep Think offers maximum parallel reasoning but remains limited to Google AI Ultra subscribers, so the deepest responses require the paid Ultra plan.
- This rollout arrives as Gemini 3.5 Flash is deployed and Gemini 3.5 Pro is due later this month, a shift that could raise compute demand, speed credit consumption for heavy users, and reshape how people choose models for complex tasks.