Google полностью переделала Gemini: уровни мышления вместо Thinking модели

Google has completely redesigned Gemini. Both the web version and the mobile app have been updated—new interface, built-in thinking depth modes, a new lightweight Flash Lite model. The update is substantial and visible. But it brought both gains and losses. Users lost the separate Thinking model and faced stricter limitations on Pro subscription.

Three Levels Instead of a Separate Thinking Model

The main change is in the architecture. Previously, Thinking was a separate model that had to be selected from the menu. Now Google has integrated the choice of thinking level directly into the main chat interface—three buttons for any question.

• Fast — instant answers, on-the-fly analysis, without deep thinking
• Balanced — standard mode, optimal for most everyday tasks
• Deep — extended thinking, for complex questions, analytics, creative tasks

It sounds more convenient. You don't need to navigate menus—just press a button in the chat and type your query. You get results faster than when searching for a separate model. But there's a catch. The Deep mode is not a full replacement for the Thinking model. It runs noticeably faster than Thinking, consumes fewer computational resources, but also analyzes less deeply. It thinks, but not for as long. For complex scientific questions, building financial models, lengthy code—Thinking was stronger. Deep falls short.

Flash Lite and the Redesigned Interface

The new Flash Lite model is designed for quick queries. Language translations, text summaries, factual information retrieval, simple questions like 'when...', 'where...', 'how much...'. It delivers results almost instantly because it's specially optimized for speed. It consumes fewer tokens, answers are shorter, but most importantly, accurate.

The entire Gemini interface has been redesigned from scratch. Unnecessary visual elements have been removed, buttons are arranged more intuitively. On mobile devices, it's now more convenient: mode selection is clearly visible without hidden menus. The desktop version has also been redesigned—cleaner, more minimalist.

Losses: Thinking Disappeared, Pro Limited

But there are serious losses too. The Thinking model has been completely removed. Those who started actively using it over the past couple of weeks suddenly found themselves without the tool. The Deep mode doesn't replace it—it works simpler and faster, but is less analytical.

Pro subscription has received stricter limitations. Previously, subscribers could send more requests per day. Now the cap is lower. Google has not disclosed exact figures, but Reddit communities and forums are full of complaints. People who pay for the product received fewer capabilities. Free users got access to modes (Fast, Balanced, Deep), but with a limit on the number of Deep queries per day. It's strange: you pay for features, you pay for restrictions, and yet you lose functionality.

What This Shows

Google is clearly changing its strategy. Instead of separate specialized models, the company is embedding options directly into the main product. It's more convenient in the interface, but offers less flexibility for experienced users who need deep tools.

Pro subscription is becoming less attractive. Restrictions are growing, functionality is becoming dull, and the best features are increasingly available in the free version. Typical commercial dynamics: first you generously offer features, then you start restricting.

For OpenAI, this is a window of opportunity. Their o1 and o3 models remain more powerful at deep analysis, and while Google checks boxes in the interface, competitors can focus on quality.