Gemini 3 hit the stage without any fanfare or long introductions, but it was clear from the get-go that Google was no longer content with chasing the leaders in the LLM race \u2014 it wanted to lead again. The model was immediately integrated into the Gemini app and the AI Search engine. Meanwhile, a more analytical version, Gemini 3 DeepThink, is waiting in the wings, set to debut after additional safety tests.<\/p>
Head of Product Tulsee Doshi talks about a \u201cmassive jump in reasoning\u201d and \u201ca level of depth and nuance that we haven’t seen before.\u201d While we often hear such claims in the AI industry, this time the facts really back up the statements.<\/p>
The power of that jump is clear in the benchmarks. Gemini 3 set a new record in Humanity’s Last Exam, scoring 37.4 points, more than any other model in history. This has been a big leap forward, as the previous leader (GPT-5 Pro) only scored 31.64. Gemini 3 also topped the LMArena ranking, which is a benchmark where users, not algorithms, decide the leader.<\/p>
On top of that, there’s something the competition doesn’t have \u2014 scale. Today, over 650 million users and 13 million developers are using Gemini. Google doesn’t need to ask people to test a new model. All they have to do is swap out the engine under the tools people use every day.<\/p>
Alongside its base model, Google has rolled out Antigravity \u2014 an innovative developer interface powered by Gemini. This tool enables agentic coding in a multi-screen environment, seamlessly integrating a chat window, a prompt line and a browser preview that shows live updates as changes are made.<\/p>
Antigravity aims to compete with solutions like Warp or Cursor 2.0. Google assures that the agent works seamlessly with editors, terminals and browsers, which is designed to streamline the app building process for developers.<\/p>
Google also announced a variant of the model called Gemini 3 DeepThink, which is more focused on research tasks. This version will be made available to subscribers of the Google AI Ultra plan in the coming weeks, after it undergoes additional rounds of safety testing.<\/p>
The launch of Gemini 3 has changed the game in how we talk about competition in the AI market today. Over the past few months, the narrative was all about OpenAI leading the way, Anthropic offering the most stable product, and Google trying to find its place. But now, things look different. Google hasn’t just caught up \u2014 it’s unveiled a model that truly pushes the boundaries in the most critical category: reasoning ability. Plus, the Antigravity tool clearly shows that Google wants to compete in the agent category too, and not just as an experiment but as a daily use solution.<\/p>
So, Gemini 3 isn’t just an update to the model \u2014 it’s a signal of a strategic shift. Google is doubling down on reasoning, agents and instant distribution of its tools. This indicates that despite the achievements of Gemini 3, this phase of the race will hinge not so much on specs, but on the model’s ability to cooperate with the user and solve real-world problems.<\/p>","protected":false},"excerpt":{"rendered":"
Gemini 3 is smashing records in reasoning tests, and soon we’ll see an even more advanced version of DeepThink. Alongside them, the Antigravity coding environment is making its debut.<\/p>\n","protected":false},"author":465,"featured_media":16205,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"rank_math_lock_modified_date":false,"footnotes":""},"categories":[813],"tags":[],"popular":[],"difficulty-level":[36],"ppma_author":[892],"class_list":["post-16246","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-news-2","difficulty-level-easy"],"acf":[],"authors":[{"term_id":892,"user_id":465,"is_guest":0,"slug":"kmironczuk","display_name":"Krzysztof Miro\u0144czuk","avatar_url":{"url":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/10\/awatar-2.png","url2x":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/10\/awatar-2.png"},"first_name":"Krzysztof","last_name":"Miro\u0144czuk","user_url":"","job_title":"","description":"Od lat zajmuj\u0119 si\u0119 nowymi technologiami w biznesie, edukacji i codziennym \u017cyciu. W centrum mojej uwagi pozostaje cz\u0142owiek \u2013 i to, by technologia wyr\u00f3wnywa\u0142a szanse, zamiast tworzy\u0107 bariery."}],"_links":{"self":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/16246","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/users\/465"}],"replies":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/comments?post=16246"}],"version-history":[{"count":1,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/16246\/revisions"}],"predecessor-version":[{"id":16247,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/16246\/revisions\/16247"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media\/16205"}],"wp:attachment":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media?parent=16246"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/categories?post=16246"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/tags?post=16246"},{"taxonomy":"popular","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/popular?post=16246"},{"taxonomy":"difficulty-level","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/difficulty-level?post=16246"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/ppma_author?post=16246"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}