Google Unveils Gemini 2.5 Deep Think: A Powerful AI Model for Complex Problem-Solving

Google releases Gemini 2.5 Deep Think for AI Ultra subscribers

Google is unleashing its most powerful Gemini model today, but you probably won't be able to try it. After revealing Gemini 2.5 Deep Think at the I/O conference back in May, Google is making this AI available in the Gemini app. Deep Think is designed for the most complex queries, which means it uses more compute resources than other models. So it should come as no surprise that only those subscribing to Google's $250 AI Ultra plan will be able to access it. Deep Think is based on the same foundation as Gemini 2.5 Pro, but it increases the "thinking time" with greater parallel analysis. According to Google, Deep Think explores multiple approaches to a problem, even revisiting and remixing the various hypotheses it generates. This process helps it create a higher-quality output. Like some other heavyweight Gemini tools, Deep Think takes several minutes to come up with an answer. This apparently makes the AI more adept at design aesthetics, scientific reasoning, and coding. Google has exposed Deep Think to the usual battery of benchmarks, showing that it surpasses the standard Gemini 2.5 Pro and competing models like OpenAI o3 and Grok 4. Deep Think shows a particularly large gain in Humanity's Last Exam, a collection of 2,500 complex, multi-modal questions that cover more than 100 subjects. Other models top out at 20 or 25 percent, but Gemini 2.5 Deep Think managed a score of 34.8 percent.

Google rolls out Gemini Deep Think AI, a reasoning model that tests multiple ideas in parallel | TechCrunch

Google DeepMind is rolling out Gemini 2.5 Deep Think, which, the company says, is its most advanced AI reasoning model, able to answer questions by exploring and considering multiple ideas simultaneously and then using those outputs to choose the best answer. Subscribers to Google's $250-per-month Ultra subscription will gain access to Gemini 2.5 Deep Think in the Gemini app starting Friday. First unveiled in May at Google I/O 2025, Gemini 2.5 Deep Think is Google's first publicly available multi-agent model. These systems spawn AI multiple agents to tackle a question in parallel, a process that uses significantly more computational resources than a single agent, but tends to result in better answers. Google used a variation of Gemini 2.5 Deep Think to score a gold medal at this year's International Math Olympiad (IMO). Alongside Gemini 2.5 Deep Think, the company says it is releasing the model it used at the IMO to a select group of mathematicians and academics. Google says this AI model "takes hours to reason," instead of seconds or minutes like most consumer-facing AI models. The company hopes the IMO model will enhance research efforts, and aims to get feedback on how to improve the multi-agent system for academic use cases. Google notes that the Gemini 2.5 Deep Think model is a significant improvement over what it announced at I/O. The company also claims to have developed "novel reinforcement learning techniques" to encourage Gemini 2.5 Deep Think to make better use of its reasoning paths. "Deep Think can help people tackle problems that require creativity, strategic planning and making improvements step-by-step," said Google in a blog post shared with TechCrunch. The company says Gemini 2.5 Deep Think achieves state-of-the-art performance on Humanity's Last Exam (HLE) -- a challenging test measuring AI's ability to answer thousands of crowdsourced questions across math, humanities, and science. Google claims its model scored 34.8% on HLE (without tools), compared to xAI's Grok 4, which scored 25.4%, and OpenAI's o3, which scored 20.3%. Google also says Gemini 2.5 Deep Think outperforms AI models from OpenAI, xAI, and Anthropic on LiveCodeBench6, a challenging test of competitive coding tasks. Google's model scored 87.6%, whereas Grok 4 scored 79%, and OpenAI's o3 scored 72%. Gemini 2.5 Deep Think automatically works with tools such as code execution and Google Search, and the company says it's capable of producing "much longer responses" than traditional AI models. In Google's testing, the model produced more detailed and aesthetically pleasing web development tasks compared to other AI models. The company claims the model could aid researchers and "potentially accelerate the path to discovery." It seems that several leading AI labs are converging around the multi-agent approach. Elon Musk's xAI recently released a multi-agent system of its own, Grok 4 Heavy, which it says was able to achieve industry leading performance on several benchmarks. OpenAI researcher Noam Brown said on a podcast that the unreleased AI model the company used to achieve a gold medal at this year's International Math Olympiad (IMO) was also a multi-agent system. Meanwhile, Anthropic's Research agent, which generates thorough research briefs, is also powered by a multi-agent system. Despite the strong performance, it seems that multi-agent systems are even costlier to serve than traditional AI models. That means tech companies may keep these systems gated behind their most expensive subscription plans, which xAI and now Google have chosen to do. In the coming weeks, Google says it plans to share Gemini 2.5 Deep Think with a select group of testers via the Gemini API. The company says it wants to better understand how developers and enterprises may use its multi-agent system.

Google's Powerful New AI Model Can Solve Your Most Complex Problems. If You Can Afford It

Expertise Artificial intelligence, home energy, heating and cooling, home technology. A supercharged version of Google's Gemini 2.5 large language model recently reached gold medal status at the International Mathematical Olympiad. Now you can ask a version of it (only a bronze medalist) to answer your toughest math questions. Like, how am I going to pay $250 a month for this AI subscription? Naturally, this new version of Google's Gemini 2.5 Deep Think is designed for complicated questions that require much more work than you'd expect from a free or cheap AI chatbot. Google said it's designed to handle prompts that require strategic planning, creativity and deep reasoning. Google's making it available now to subscribers of the company's $250-per-month AI Ultra plan. That subscription, first announced earlier this year at Google's I/O developer conference, is a steep price to pay for generative AI tools. The tech giant teased then that this model would be among the offerings to make that price worthwhile. (You can still access other Gemini models either for free or at a $20-per-month tier.) Generative AI developers are intent on building ever more powerful models, and one big way they're doing so is by expanding their ability for research and increasing the amount of time these models can work. OpenAI has a Deep Research tool for ChatGPT users, and Google's Gemini has a tool by the same name for all users. While those research modes allow models more time, Deep Think uses "parallel thinking" to consider different ideas at once, combining or revising them as it goes. Deep Think also just has more time to think and process, which helps it find better answers but also costs more money and uses more computing power.

Google releases its award-winning Math Olympiad model, but it'll come at a price

Two weeks ago, Google and OpenAI touted their models' award-winning performance at the International Math Olympiad (IMO). Now, Google is making a version of its model available to the public. Also: This one feature could make GPT-5 a true game changer (if OpenAI gets it right) On Friday, Google launched Deep Think in the Gemini app for Google Ultra subscribers, a premium subscription tier that costs $250 per year or $125 for the first three months. Although the model is a variation of the one that achieved the gold-medal standard at IMO, it is faster for everyday tasks. Internal evaluations suggested the model reaches bronze-level on the 2025 IMO benchmark. The superior performance solving complex problems is enabled by a parallel thinking technique, which allows the model to simultaneously generate and process multiple ideas, even combining different ones as necessary to find the best answer. Also: Google upgrades AI Mode with Canvas and 3 other new features - how to try them Other factors contributing to the high performance include an extended inference time, also known as thinking time, which allows Deep Think to explore more options before arriving at an answer, and new reinforcement learning techniques that help the model to become a better problem-solver over time. According to Google, Deep Think excels at iterative development and design, as seen in the image above, scientific and mathematical discovery, and coding. These results are reflected in Gemini 2.5 Deep Think's performance across state-of-the-art benchmarks, including Humanity's Last Exam, an exam with multi-modal questions in over 100 subjects, such as math, science, and the humanities. Google also shared that the Gemini 2.5 Deep Think has shown better content safety and tone-objectivity compared to Gemini 2.5 Pro, with the caveat that it denied benign requests at a higher rate. Google AI Ultra subscribers can access Deep Think in the Gemini app with a fixed set of prompts daily. To select the model, toggle "Deep Think" in the prompt bar when selecting 2.5 Pro on the model selector. The company also shared that it's working to release Deep Think, with and without tools, to a set of trusted testers via the Gemini API in the coming weeks. Also: OpenAI teases imminent GPT-5 launch. Here's what to expect The Gemini 2.5 Deep Think model that achieved the gold-medal standard will be shared with a small group of mathematicians and academics. The intention is that this model will be used to advance their work, and it is hoped the lessons will provide feedback for improvements.

You can use Google's Math Olympiad-winning Deep Think AI model now - for a price

Two weeks ago, Google and OpenAI touted their models' award-winning performance at the International Math Olympiad (IMO). Now, Google is making a version of its model available to the public. Also: This one feature could make GPT-5 a true game changer (if OpenAI gets it right) On Friday, Google launched Deep Think in the Gemini app for Google Ultra subscribers, a premium subscription tier that costs $250 per year or $125 for the first three months. Although the model is a variation of the one that achieved the gold-medal standard at IMO, it is faster for everyday tasks. Internal evaluations suggested the model reaches bronze-level on the 2025 IMO benchmark. The superior performance solving complex problems is enabled by a parallel thinking technique, which allows the model to simultaneously generate and process multiple ideas, even combining different ones as necessary to find the best answer. Also: Google upgrades AI Mode with Canvas and 3 other new features - how to try them Other factors contributing to the high performance include an extended inference time, also known as thinking time, which allows Deep Think to explore more options before arriving at an answer, and new reinforcement learning techniques that help the model to become a better problem-solver over time. According to Google, Deep Think excels at iterative development and design, as seen in the image above, scientific and mathematical discovery, and coding. These results are reflected in Gemini 2.5 Deep Think's performance across state-of-the-art benchmarks, including Humanity's Last Exam, an exam with multi-modal questions in over 100 subjects, such as math, science, and the humanities. Google also shared that the Gemini 2.5 Deep Think has shown better content safety and tone-objectivity compared to Gemini 2.5 Pro, with the caveat that it denied benign requests at a higher rate. Google AI Ultra subscribers can access Deep Think in the Gemini app with a fixed set of prompts daily. To select the model, toggle "Deep Think" in the prompt bar when selecting 2.5 Pro on the model selector. The company also shared that it's working to release Deep Think, with and without tools, to a set of trusted testers via the Gemini API in the coming weeks. Also: OpenAI teases imminent GPT-5 launch. Here's what to expect The Gemini 2.5 Deep Think model that achieved the gold-medal standard will be shared with a small group of mathematicians and academics. The intention is that this model will be used to advance their work, and it is hoped the lessons will provide feedback for improvements.

Gemini adds powerful new Deep Think model - what it does and who can try it

Two weeks ago, Google and OpenAI touted their models' award-winning performance at the International Math Olympiad (IMO). Now, Google is making a version of its model available to the public. Also: This one feature could make GPT-5 a true game changer (if OpenAI gets it right) On Friday, Google launched Deep Think in the Gemini app for Google Ultra subscribers, a premium subscription tier that costs $250 per year or $125 for the first three months. Although the model is a variation of the one that achieved the gold-medal standard at IMO, it is faster for everyday tasks. Internal evaluations suggested the model reaches bronze-level on the 2025 IMO benchmark. The superior performance solving complex problems is enabled by a parallel thinking technique, which allows the model to simultaneously generate and process multiple ideas, even combining different ones as necessary to find the best answer. Also: Google upgrades AI Mode with Canvas and 3 other new features - how to try them Other factors contributing to the high performance include an extended inference time, also known as thinking time, which allows Deep Think to explore more options before arriving at an answer, and new reinforcement learning techniques that help the model to become a better problem-solver over time. According to Google, Deep Think excels at iterative development and design, as seen in the image above, scientific and mathematical discovery, and coding. These results are reflected in Gemini 2.5 Deep Think's performance across state-of-the-art benchmarks, including Humanity's Last Exam, an exam with multi-modal questions in over 100 subjects, such as math, science, and the humanities. Google also shared that the Gemini 2.5 Deep Think has shown better content safety and tone-objectivity compared to Gemini 2.5 Pro, with the caveat that it denied benign requests at a higher rate. Google AI Ultra subscribers can access Deep Think in the Gemini app with a fixed set of prompts daily. To select the model, toggle "Deep Think" in the prompt bar when selecting 2.5 Pro on the model selector. The company also shared that it's working to release Deep Think, with and without tools, to a set of trusted testers via the Gemini API in the coming weeks. Also: OpenAI teases imminent GPT-5 launch. Here's what to expect The Gemini 2.5 Deep Think model that achieved the gold-medal standard will be shared with a small group of mathematicians and academics. The intention is that this model will be used to advance their work, and it is hoped the lessons will provide feedback for improvements.

Google rolls out its most powerful Gemini model yet

Google AI Ultra subscribers to Deep Think, Google's most advanced reasoning model yet for Gemini 2.5. The new model was first unveiled at , and after feedback from "early trusted testers," is now being rolled out to a wider audience. For now, access is limited to the Google offers, which will run you a cool $250 per month. The public version of Deep Think is a variation of the model that recently achieved a gold-medal standard at the . Google the version available today would achieve a bronze-medal standard at the same competition according to internal testing, and is faster and better suited for daily use. The gold-medal model will still be tested further and is being shared with a small group of mathematicians and academics for research. Deep Think works by leveraging what Google describes as parallel thinking techniques. This approach allows Gemini to generate multiple ideas at once and consider them simultaneously. Deep Think is also , meaning it can take in different types of data including text, images and sound. Deep Think was tested against various AI benchmarks, such as and and outscored , and Gemini 2.5 Pro, according to Google. AI Ultra subscribers can begin using Deep Think within the today, though they will be limited to a fixed number of prompts per day. Google did not share exactly how many prompts will be allowed.

Google's Advanced AI Model Is Now Available to Tryâ€"for $250 a Month

Gemini 2.5 Deep Think uses multiple agents paths to crack complex problems Google just made one of its most advanced AI reasoning models available to the public. Like other advanced models, Gemini 2.5 Deep Think uses more than one AI agent to brainstorm answers in order to boost accuracy and result in more creative solutions. Google says the model outperformed several of its competitors in key AI benchmark tests and a variation of the model even achieved the gold-medal standard at this year’s International Mathematical Olympiad (IMO), solving five out of the six IMO problems perfectly. That research model took hours to produce solutions, but the version available today is designed for everyday use working much faster, while still achieving bronze-level IMO performance. However, in order to try out the new model users will need to hand over $250 a month for a Google AI Ultra subscription.Starting today, subscribers get access to a fixed set of prompts for the new model. They can enable “Deep Think†by toggling it in the prompt bar when selecting Gemini 2.5 Pro from the model dropdown menu in the Gemini app. Google first previewed Gemini 2.5 Deep Think back in May at its I/O developer conference, but the company says the version released today is a “significant improvement,†thanks to tester feedback and key benchmark improvements. Google says Deep Think uses parallel thinking techniques to tackle complex problems like a human would by weighing different angles and potential solutions. “This approach lets Gemini generate many ideas at once and consider them simultaneously, even revising or combining different ideas over time, before arriving at the best answer,†the company said in a blog post. Additionally, Google says it developed new reinforcement learning techniques that push the model to explore extended reasoning paths, helping Deep Think become a stronger and more intuitive problem-solver over time. This, Google claims, makes the model particularly useful for things like coding, web development, and scientific research. According to Google, Gemini 2.5 Deep Think outperformed rival models on the Humanity’s Last Exam (HLE), a 2,500-question expertise benchmark spanning subjects like math, science, and the humanities. The model achieved a score of 34.8% on the test, compared with OpenAI o3’s 20.3% and Grok 4’s 25.4% scores. Google also said it will share the gold-medal version of Gemini 2.5 Deep Think with a small group of mathematicians and academics, hoping to see how it might aid their research. The company plans to use that feedback to refine future versions of the model.

Gemini's Deep Think lets you solve much more complex problems

I finally found the perfect note-taking app after years of switching back and forth It seems like every day, Google is adding new AI tools to its apps and services. That's to be expected when it's invested so much, and that appears to be the focus of future technologies going forward. With that said, Google has announced that Google AI Ultra subscribers will now have access to even more powerful AI tools, with access to Deep Think through the Gemini app. If you need to solve complex problems, Deep Think is going to be the solution Now, most people probably aren't going to utilize the full power of Deep Think, but it's good to know that Google is providing such a service with its latest addition for AI Ultra subscribers. Furthermore, Google is sharing this tool with an exclusive group of folks in academia, where they will be able to use the tool in order to potentially further their own research. What makes Deep Think special is that it can consider many different paths when it comes to a problem and providing a solution. Google highlights its ability to utilize "parallel thinking techniques" to come up with solutions at a faster pace than what's normally possible. And since these ideas work in parallel, Deep Think can borrow from each idea, or even remix processes in order to create one solution. It's actually quite impressive once you see it in action, and will no doubt be even more impressive in capable hands. This could make breakthroughs occur even faster, and could really take the hard work out of new discoveries. Whether this is a good thing will be something that everyone can decide on their own. It probably doesn't need to be said, but this is the most advanced form of AI that Google has produced, and it far surpasses what's currently available from its other AI models, like Gemini 2.5 Pro. If you're curious about the frame of Gemini 2.5 Deep Think, you can get all the details from Google's model card. For those that are eligible, you can now utilize Deep Think within the Gemini app for Android. You'll need to activate the feature in the prompt bar, but once that's set, you'll be good to go. Naturally, Deep Think will work with other Google apps and tools in order to provide a complete solution when needed.

Gemini 2.5 Deep Think rolling out now for Google AI Ultra

After previewing at I/O 2025 in May, Google is now rolling out the Gemini 2.5 Deep Think mode to AI Ultra subscribers. Google says "Deep Think pushes the frontier of thinking capabilities by leveraging parallel thinking techniques." This approach enables Gemini to generate many ideas at once and consider them simultaneously, even revising or combining different ideas over time, before arriving at the best answer. Gemini is given more inference or "thinking time" to "explore different hypotheses, and arrive at creative solutions to complex problems." Google "developed novel reinforcement learning techniques that encourage the model to make use of these extended reasoning paths." ...thus enabling Deep Think to become a better, more intuitive problem-solver over time. Deep Think can produce "much longer responses" and leverages code execution, Google Search, and other tools. Today's release offers a "significant improvement over what was first announced" in May. Google credits feedback from early trusted testers and "research breakthroughs" for these "key benchmark improvements" across coding, science, knowledge, and reasoning. Gemini 2.5 Deep Think now scores 87.6% on LiveCodeBench (competition-level coding) versus 80.4% in May. In terms of utility, Deep Think excels at iterative development and design of "tasks that require building something complex, piece by piece." ...we've observed Deep Think can improve both the aesthetics and functionality of web development tasks. It performs well at "tough coding problems where problem formulation and careful consideration of tradeoffs and time complexity is paramount." Google sees Deep Think as being a powerful tool for highly complex scientific and mathematical problems: "It can help formulate and explore mathematical conjectures or reason through complex scientific literature, potentially accelerating the path to discovery. " For example, this Deep Think release can score Bronze-level performance in the 2025 International Mathematical Olympiad (IMO) benchmark and practically be used day-to-day. In comparison, the "variation" of the Deep Think model (which Google is now releasing to some academics) that was Gold-medal level took hours. If you're paying for Google AI Ultra ($249.99 per month in the US), Deep Think is enabled from a new button in the prompt bar when using Gemini 2.5 Pro. Subscribers will get "a few prompts" per day on mobile and web. In the coming weeks, Google will offer a with and without tools version to trusted testers via the Gemini API to explore developer and enterprise use cases.

Google's new Deep Think feature is here -- what it does and why it might not stay Ultra-exclusive for long

Google is notorious for giving away access to its best models Google just dropped one of its most ambitious AI upgrades yet. The company is rolling out Deep Think, a powerful new mode inside the Gemini app that boosts the reasoning power of its top-tier model, Gemini 2.5 Ultra. Starting today, Deep Think is available to Google AI Ultra subscribers, with a broader API release for trusted testers expected in the coming weeks. Google says Deep Think excels at challenges that require multi-step reasoning, strategy and iteration. Think of it as giving Gemini more thinking time before the AI responds. Now, instead of rushing to deliver an answer, Gemini engages in parallel thinking, a process that allows it to generate multiple ideas, evaluate them simultaneously and refine its approach before responding. It's a move that brings Gemini one step closer to emulating human-style problem-solving, whether it's debugging code, website building or tackling advanced math. Internally, the model has reached bronze-level performance on the 2025 International Mathematical Olympiad (IMO) benchmark and outperformed competitors on difficult reasoning tests like LiveCodeBench V6 and Humanity's Last Exam, which measure capabilities in science, coding and complex logic. The company also tested Deep Think with academic researchers, including mathematician Michel van Garrel, to explore its potential for tackling theoretical problems and advancing scientific discovery. While the original research model behind Deep Think took hours to solve Olympiad-level math problems, the version launching today is optimized for faster, more practical day-to-day use. To try it out, Ultra subscribers can simply toggle Deep Think on in the Gemini app when using Gemini 2.5 Pro. The mode supports longer, more detailed responses and works with tools like Google Search and code execution, though usage is currently limited to a fixed number of prompts per day. As with previous Gemini updates, Google is placing a strong emphasis on safety. The company reports that Deep Think shows improved tone and content moderation over Gemini 2.5 Pro, though it may sometimes refuse benign requests. Google says it's actively performing frontier safety evaluations and working on mitigations as model capabilities advance. For Ultra subscribers, this new feature should support app building, exploring a scientific theory and many more tasks that require a more thoughtful response. While Deep Think is currently limited to $249/month Gemini Ultra subscribers, there's a good chance it could become more widely accessible in the future. Google has a history of offering some of its most powerful AI features, like Gemini Pro and image generation, to free-tier users after initial testing. If Deep Think proves successful and stable, it's likely that Google will broaden access to showcase its capabilities and stay competitive alongside other tech giants.

Deep Think is available in the Google Gemini App. How to try it.

Google is rolling out Gemini 2.5 Deep Think in the Gemini app, bringing its AI reasoning model to Google AI Ultra subscribers. Subscribers to the AI subscription, which costs $250 per month, will be granted access to Deep Think starting Friday, August 1. Deep Think was first revealed at Google I/O 2025, where the tech giant introduced the model designed for complex math and coding. The AI reasoning model is reportedly Google's most advanced model so far, with higher reasoning capabilities and the ability to solve complex problems using parallel thinking techniques. In its official announcement, Google said that Deep Think provides "more detailed, creative and thoughtful responses" to prompts. The company says that Gemini 2.5 Deep Think outperforms competing AI models, using scores from Humanity's Last Exam (HLE), a complex AI test, as a benchmark. According to Google, its model scored 34.8% on the test, which features questions about math, sciences, and humanities. On the other hand, xAI's Grok 4 scored 25.4% and OpenAI's o3 scored 20.3%. Google's model also outperformed AI models from xAI and Anthropic on LiveCodeBench6, a test consisting of coding tasks. Google also announced that it is releasing the official version of its Gemini 2.5 Deep Think model, which achieved a gold-medal standard at the International Mathematical Olympiad (IMO) earlier in July. The company will offer this version to a select group of "trusted testers", which includes mathematicians and academics. The president of IMO called Deep Think's solutions "astonishing in many respects."

Google releases Olympiad medal-winning Gemini 2.5 'Deep Think' AI publicly -- but there's a catch...

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now Google has officially launched Gemini 2.5 Deep Think, a new variation of its AI model engineered for deeper reasoning and complex problem-solving, which made headlines last month for winning a gold medal at the International Mathematical Olympiad (IMO) -- the first time an AI model achieved the feat. However, this is unfortunately not the identical gold medal-winning model. It is in fact, a less powerful "bronze" version according to Google's blog post and Logan Kilpatrick, Product Lead for Google AI Studio. As Kilpatrick posted on the social network X: "This is a variation of our IMO gold model that is faster and more optimized for daily use. We are also giving the IMO gold full model to a set of mathematicians to test the value of the full capabilities." Now available through the Gemini mobile app, this bronze model is accessible to subscribers of Google's most expensive individual AI plan, AI Ultra, which costs $249.99 per month with a 3-month starting promotion at a reduced rate of $124.99/month for new subscribers. Google also said in its release blog post that it would bring Deep Think with and without tool usage integrations to "trusted testers" through the Gemini application programming interface (API) "in the coming weeks." Why 'Deep Think' is so powerful Gemini 2.5 Deep Think builds on the Gemini family of large language models (LLMs), adding new capabilities aimed at reasoning through sophisticated problems. It employs "parallel thinking" techniques to explore multiple ideas simultaneously and includes reinforcement learning to strengthen its step-by-step problem-solving ability over time. The model is designed for use cases that benefit from extended deliberation, such as mathematical conjecture testing, scientific research, algorithm design, and creative iteration tasks like code and design refinement. Early testers, including mathematicians such as Michel van Garrel, have used it to probe unsolved problems and generate potential proofs. AI power user and expert Ethan Mollick, a professor of the Wharton School of Business at the University of Pennsylvania, also posted on X that it was able to take a prompt he often uses to test the capabilities of new models -- "create something I can paste into p5js that will startle me with its cleverness in creating something that invokes the control panel of a starship in the distant future" -- and turned it into a 3D graphic, which is the first time any model has done that. Performance benchmarks and use cases Google highlights several key application areas for Deep Think: * Mathematics and science: The model can simulate reasoning for complex proofs, explore conjectures, and interpret dense scientific literature * Coding and algorithm design: It performs well on tasks involving performance tradeoffs, time complexity, and multi-step logic * Creative development: In design scenarios such as voxel art or user interface builds, Deep Think demonstrates stronger iterative improvement and detail enhancement The model also leads performance in benchmark evaluations such as LiveCodeBench V6 (for coding ability) and Humanity's Last Exam (covering math, science, and reasoning). It outscored Gemini 2.5 Pro and competing models like OpenAI's GPT-4 and xAI's Grok 4 by double digit margins on some categories (Reasoning & Knowledge, Code generation, and IMO 2025 Mathematics). Gemini 2.5 Deep Think vs. Gemini 2.5 Pro While both Deep Think and Gemini 2.5 Pro are part of the Gemini 2.5 model family, Google positions Deep Think as a more capable and analytically skilled variant, particularly when it comes to complex reasoning and multi-step problem-solving. This improvement stems from the use of parallel thinking and reinforcement learning techniques, which enable the model to simulate deeper cognitive deliberation. In its official communication, Google describes Deep Think as better at handling nuanced prompts, exploring multiple hypotheses, and producing more refined outputs. This is supported by side-by-side comparisons in voxel art generation, where Deep Think adds more texture, structural fidelity, and compositional diversity than 2.5 Pro. The improvements aren't just visual or anecdotal. Google reports that Deep Think outperforms Gemini 2.5 Pro on multiple technical benchmarks related to reasoning, code generation, and cross-domain expertise. However, these gains come with tradeoffs in responsiveness and prompt acceptance. Here's a breakdown: Google notes that Deep Think's higher refusal rate is an area of active investigation. This may limit its flexibility in handling ambiguous or informal queries compared to 2.5 Pro. In contrast, 2.5 Pro remains better suited for users who prioritize speed and responsiveness, especially for lighter, general-purpose tasks. This differentiation allows users to choose based on their priorities: 2.5 Pro for speed and fluidity, or Deep Think for rigor and reflection. Not the gold medal winning model, just a bronze In July, Google DeepMind made headlines when a more advanced version of the Gemini Deep Think model achieved official gold-medal status at the 2025 IMO -- the world's most prestigious mathematics competition for high school students. The system solved five of six challenging problems and became the first AI to receive gold-level scoring from the IMO. Demis Hassabis, CEO of Google DeepMind, announced the achievement on X, stating the model had solved problems end-to-end in natural language -- without needing translation into formal programming syntax. The IMO board confirmed the model scored 35 out of a possible 42 points, well above the gold threshold. Gemini 2.5 Deep Think's solutions were described by competition president Gregor Dolinar as clear, precise, and in many cases, easier to follow than those of human competitors. However, the Gemini 2.5 Deep Think released to users is not that same competition model, rather, a lower performing but apparently faster version. How to access Deep Think now Gemini 2.5 Deep Think is available exclusively on the Google Gemini mobile app for iOS and Android at this time to users on the Google AI Ultra plan, part of the Google One subscription lineup, with pricing as follows. * Promotional offer: $124.99/month for 3 months, then it kicks up to... * Standard rate: $249.99/month * Included features: 30 TB of storage, access to the Gemini app with Deep Think and Veo 3, as well as tools like Flow, Whisk, and 12,500 monthly AI credits Subscribers can activate Deep Think in the Gemini app by selecting the 2.5 Pro model and toggling the "Deep Think" option. It supports a fixed number of prompts per day and is integrated with capabilities like code execution and Google Search. The model also generates longer and more detailed outputs compared to standard versions. The lower-tier Google AI Pro plan, priced at $19.99/month (with a free trial), does not include access to Deep Think, nor does the free Gemini AI service. Why it matters for enterprise technical decision-makers Gemini 2.5 Deep Think represents the practical application of a major research milestone. It allows enterprises and organizations to tap into a Math Olympiad medal-winning model and have it join their staff, albeit only through an individual user account now. For researchers receiving the full IMO-grade model, it offers a glimpse into the future of collaborative AI in mathematics. For Ultra subscribers, Deep Think provides a powerful step toward more capable and context-aware AI assistance, now running in the palm of their hand.

Try Deep Think in the Gemini app

Today, we're making Deep Think available in the Gemini app to Google AI Ultra subscribers - the latest in a lineup of extremely capable AI tools and features made exclusively available to them. This new release incorporates feedback from early trusted testers and research breakthroughs. It's a significant improvement over what was first announced at I/O, as measured in terms of key benchmark improvements and trusted tester feedback. It is a variation of the model that recently achieved the gold-medal standard at this year's International Mathematical Olympiad (IMO). While that model takes hours to reason about complex math problems, today's release is faster and more usable day-to-day, while still reaching Bronze-level performance on the 2025 IMO benchmark, based on internal evaluations. Deep Think could be a powerful tool in creative problem solving:

Google rolls out powerful creative problem-solving AI model Deep Think to the Gemini app - SiliconANGLE

Google rolls out powerful creative problem-solving AI model Deep Think to the Gemini app Google DeepMind, Alphabet Inc.'s artificial intelligence research arm, today announced the rollout of Gemini 2.5 Deep Think, a new creative problem-solving AI model. The company stated the model is capable of addressing complex problems by considering multiple ideas simultaneously and choosing the best answer. The new tool will be available on the Gemini app for Google AI Ultra subscribers for $250 per month. Google said model extends the Gemini model's parallel "thinking time." Similar to how people explore complex problems by examining different angles, Deep Think works through problems by generating and simultaneously considering multiple variations. It can also combine different ideas over time before arriving at a conclusion. "By extending the inference time, or 'thinking time,' we give Gemini more time to explore different hypotheses and arrive at creative solutions to complex problems," the Deep Think team said. The team also developed new reinforcement learning techniques, a type of training model for AI, to encourage the model to carefully consider problems. This enables Deep Think to be a more intuitive problem-solver, the team added. Deep Think can be used for iterative development and design, such as for creative projects or complex development tasks. The team noted that it also excels at research and mathematical study, making it an excellent tool for science, and is very good at tough coding problems where problem formulation can be a sticky subject. "We've been impressed by Deep Think's performance on tasks that require building something complex, piece by piece," the Deep Think team said. "For example, we've observed Deep Think can improve both the aesthetics and functionality of web development tasks." The new model's release comes after an earlier version achieved the gold-medal standard at this year's International Mathematical Olympiad. While that version took hours to reason through complex math problems, today's release is designed to be more usable for everyday creative thinking while providing Bronze-level performance. The team said Deep Think's performance stands out against other models in challenging benchmarks that measure coding, science knowledge, and reasoning capabilities. For example, without tool use, Gemini 2.5 Deep Think excelled in LiveCodeBench V6, which measures code performance, and Humanity's Last Exam, a benchmark that challenges expertise across a broad variety of knowledge domains such as science and math. Gemini app users with access to Deep Think can toggle it on within the prompt bar when selecting 2.5 Pro in the model dropdown. It automatically works with tools like code execution and Google Search and can produce much longer responses. The company said it is working to release Deep Think, with and without tools, to trusted testers via the Gemini application programming interface in the coming weeks.

Google Launches Gemini 2.5 Deep Think, Outperforms Grok-4 & OpenAI o3 | AIM

Google also plans to expand access through the Gemini API, allowing developers and enterprise testers to explore its applications more broadly. Google has launched Deep Think, a new AI reasoning feature, in the Gemini app for Google AI Ultra subscribers. The model, based on Gemini 2.5, is designed to handle complex reasoning tasks and is now available for broader use after months of internal testing and research improvements. Subscribers can enable Deep Think in the Gemini app by selecting Gemini 2.5 Pro and toggling the "Deep Think" option. The feature comes with a fixed number of prompts per day and works with tools like code execution and Google Search. Google also plans to expand access through the Gemini API, allowing developers and enterprise testers to explore its applications more broadly. https://twitter.com/GoogleDeepMind/status/1951239132950204439?t=-wK49ZnJexRfVAI_tc_-jQ&s=03 Alongside the public

Google launches Deep Think for $250/month AI tier

Google has launched Deep Think, its most advanced Gemini reasoning model, exclusively for AI Ultra subscribers at $250 monthly. This rollout follows initial testing with select users. Deep Think debuted at Google I/O earlier this year. After incorporating feedback from early trusted testers, the model is now accessible to all AI Ultra tier subscribers. Google's $250 monthly subscription represents its highest-priced AI access tier. The public release constitutes a modified version of a model that attained gold-medal level performance at the International Mathematical Olympiad. Internal evaluations indicate this iteration achieves a bronze-medal standard at the same competition. Google optimized this variant for daily usability and processing speed. The original gold-medal model remains under evaluation with mathematicians and academics for research purposes. Google employs parallel thinking techniques within Deep Think, enabling simultaneous generation and consideration of multiple ideas. The model processes multimodal inputs, including text, images, and audio. In benchmark assessments -- specifically Humanity's Last Exam and LiveCodeBench -- Deep Think surpassed scores from OpenAI's o3, xAI's Grok 4, and Google's own Gemini 2.5 Pro, according to company testing data. AI Ultra members can immediately utilize Deep Think via the Gemini application. Google confirmed a daily cap on prompt interactions exists but did not disclose the specific numerical limit.

Award-winning variant of Gemini's AI model is live, confirms CEO Sundar Pichai - The Economic Times

The Gemini 2.5 Deep Think model has been rolled out to a small group of mathematicians and academics to take early feedback and enhance the model's research and inquiry.Google has rolled out a version of its artificial intelligence (AI) model Deep Think in the Gemini app, its chief executive, Sundar Pichai, said in a post on X. Gemini 2.5 Deep Think achieved the gold-medal standard at the International Math Olympiad (IMO) competition held between July 10-20 on Australia's Sunshine Coast. The results marked the first time that AI systems such as OpenAI and Google's Gemini crossed the gold-medal scoring threshold at the IMO for high school students. However, when it comes to solving complex math problems, the current model takes hours to reason and solve user queries. In a blog post, the company said the currently available model is capable of achieving bronze-level performance on the 2025 IMO benchmark, based on internal evaluations. Additionally, the company said, the Gemini 2.5 Deep Think model has been rolled out to a small group of mathematicians and academics to take early feedback and enhance the model's research and inquiry. Parallel thinking techniques Google's Deep Think model utilises parallel thinking techniques, an approach that lets Gemini generate multiple ideas at once, weighing potential solutions to solve complex problems and considering them simultaneously before giving the final answer. The idea is to integrate creative thinking capabilities into the AI model by exploring different hypotheses and giving it "thinking time." Gemini positions Deep Think as a powerful tool for researchers, as it can reason through "complex scientific literature" to highly complex mathematical problems and along with coding. State-of-the-art performance Gemini 2.5 Deep Think achieved state-of-the-art performance across benchmarks like LiveCodeBench V6, which measures competitive code performance, and humanity's last exam, which measures expertise in different domains, including science and math. AI taking charge According to a Reuters report published on July 22, Junehyuk Jung, a math professor at Brown University and visiting researcher in Google's DeepMind AI unit, said AI is less than a year away from being used by mathematicians to crack unsolved research problems at the frontier of the field. CEO Sundar Pichai had also mentioned that AI has brought about a significant platform shift, allowing people, businesses, and communities all over the world to access decades of research.

Gemini's gold-medal AI model is live, announces Sundar Pichai - The Economic Times

The Gemini 2.5 Deep Think model has been rolled out to a small group of mathematicians and academics to take early feedback and enhance the model's research and inquiry.Google has rolled out a version of its artificial intelligence (AI) model Deep Think in the Gemini app, its chief executive, Sundar Pichai, said in a post on X. Gemini 2.5 Deep Think achieved the gold-medal standard at the International Math Olympiad (IMO) competition held between July 10-20 on Australia's Sunshine Coast. The results marked the first time that AI systems such as OpenAI and Google's Gemini crossed the gold-medal scoring threshold at the IMO for high school students. However, when it comes to solving complex math problems, the current model takes hours to reason and solve user queries. In a blog post, the company said the currently available model is capable of achieving bronze-level performance on the 2025 IMO benchmark, based on internal evaluations. Additionally, the company said, the Gemini 2.5 Deep Think model has been rolled out to a small group of mathematicians and academics to take early feedback and enhance the model's research and inquiry. Parallel thinking techniques Google's Deep Think model utilises parallel thinking techniques, an approach that lets Gemini generate multiple ideas at once, weighing potential solutions to solve complex problems and considering them simultaneously before giving the final answer. The idea is to integrate creative thinking capabilities into the AI model by exploring different hypotheses and giving it "thinking time." Gemini positions Deep Think as a powerful tool for researchers, as it can reason through "complex scientific literature" to highly complex mathematical problems and along with coding. State-of-the-art performance Gemini 2.5 Deep Think achieved state-of-the-art performance across benchmarks like LiveCodeBench V6, which measures competitive code performance, and humanity's last exam, which measures expertise in different domains, including science and math. AI taking charge According to a Reuters report published on July 22, Junehyuk Jung, a math professor at Brown University and visiting researcher in Google's DeepMind AI unit, said AI is less than a year away from being used by mathematicians to crack unsolved research problems at the frontier of the field. CEO Sundar Pichai had also mentioned that AI has brought about a significant platform shift, allowing people, businesses, and communities all over the world to access decades of research.

You Can Now Use Gemini 2.5 Deep Think AI That Can Win a Math Olympiad

It's rolling out to Google AI Ultra subscribers in the Gemini app. Google has finally released the Gemini 2.5 Deep Think mode, which has delivered breakthrough performance in key areas like mathematical research, coding, and scientific reasoning. This is a faster variation of the same AI model that recently achieved the Gold medal at the IMO 2025. The IMO variant takes hours to reason on complex problems, so Google has developed a faster version for day-to-day usage, and it's being released as Gemini 2.5 Deep Think. Even then, this version achieves the Bronze medal on the IMO 2025 benchmark, according to internal evaluation. The new Gemini 2.5 Deep Think mode utilizes new methods such as parallel thinking to explore many ideas at once, extended thinking time to reason through complex problems, and novel reinforcement learning (RL) techniques to "make use of these extended reasoning paths." These are new research ideas in the AI field, and both Google DeepMind and OpenAI are working on parallel test-time compute. This approach is yielding great results in generating mathematical conjectures, creative problem solving, coding challenges, understanding scientific literature, and more. In the Humanity's Last Exam benchmark, Gemini 2.5 Deep Think scored the highest, achieving 34.8% and outranking OpenAI o3 and Grok 4. Similarly, in LiveCodeBench, Gemini 2.5 Deep Think does much better than frontier AI models from OpenAI and xAI. Google says Gemini 2.5 Deep Think is rolling out to Google AI Ultra subscribers in the Gemini app, those who pay $250 per month. You can choose the Gemini 2.5 Pro model and select the "Deep Think" toggle to activate it. Meanwhile, Google is also releasing the full version of Gemini 2.5 Deep Think to select mathematicians.

Gemini Deep Think : Solving Complex Applications in Math and Beyond

What if a machine could think as deeply as a human mathematician, solving problems so intricate they stump even the brightest minds? Enter Gemini Deep Think, an advanced AI model that has not only redefined what artificial intelligence can achieve but also challenged our understanding of reasoning itself. With its new performance at the International Mathematical Olympiad (IMO) -- a stage traditionally dominated by human brilliance -- this AI has proven it can rival the sharpest intellects in tackling complex algebra, geometry, and number theory. Yet, this achievement raises a pressing question: can such computational power ever balance its brilliance with real-world practicality? In this overview Sam Witteveen explores how Gemini Deep Think is reshaping the boundaries of AI reasoning, from its innovative use of parallel reasoning chains to its potential applications in fields like 3D modeling and algorithm design. But this isn't just a story of triumph; it's also one of trade-offs. While the model's ability to solve intricate problems with precision is unparalleled, its high computational demands and extended processing times reveal the challenges of scaling such technology. As we delve deeper, you'll discover not only the promise of this AI marvel but also the hurdles it must overcome to truly transform industries and redefine intelligence itself. What does this mean for the future of human and machine collaboration? Let's explore. The International Mathematical Olympiad (IMO) is widely regarded as one of the most prestigious global competitions, challenging high school students to solve intricate problems in algebra, geometry, and number theory. For the first time in history, an AI model -- Gemini Deep Think -- has matched the performance of top human participants, scoring an impressive 35 out of 42 points. This achievement is a testament to the model's ability to engage in logical problem-solving and advanced mathematical reasoning, areas traditionally dominated by human intelligence. By excelling in such a rigorous competition, Gemini Deep Think has not only proven its technical capabilities but also highlighted the potential for AI to complement human expertise in solving complex problems. This milestone reflects a significant step forward in AI's evolution, showcasing its capacity to operate in domains that require deep analytical thinking. Gemini Deep Think represents a significant advancement in AI reasoning by introducing innovative methodologies that set it apart from earlier models. One of its most notable features is the use of parallel reasoning chains, which allow the model to evaluate multiple solutions simultaneously and select the most effective one. This capability enables it to excel in tasks such as solving complex algebraic equations, generating structured outputs like 3D models, and addressing intricate coding challenges. The model's advanced reasoning capabilities, however, come with a trade-off. Solving complex problems can take between 10 to 20 minutes, reflecting its substantial computational demands. While this processing time underscores the sophistication of its algorithms, it also highlights the need for optimization to improve efficiency. The balance between computational power and practical usability remains a key area for development as AI continues to evolve. Check out more relevant guides from our extensive collection on advanced AI models that you might find useful. Gemini Deep Think builds upon and surpasses the capabilities of its predecessors, such as AlphaProof and AlphaGeometry. Unlike these earlier models, which relied heavily on specialized mathematical languages like Lean, Gemini Deep Think processes problems directly, offering greater flexibility and adaptability. This advancement allows it to handle a broader range of tasks, from solving mathematical benchmarks to tackling logical reasoning challenges across diverse domains. Despite its superior performance, the model's computational intensity presents a significant limitation. Its extended processing times make it less practical for applications where speed is critical, such as real-time decision-making or dynamic problem-solving environments. Addressing these limitations will be essential for making sure its broader applicability and integration into various industries. The versatility of Gemini Deep Think opens up a wide range of potential applications across multiple fields. Some of the most promising use cases include: However, the model's limitations cannot be ignored. Its long processing times and high computational requirements pose challenges for industries that rely on rapid decision-making or real-time solutions. These constraints highlight the need for further refinement to make the model more practical and accessible for real-world applications. Without addressing these issues, its adoption may remain limited to niche areas where processing time is less critical. As AI technology continues to advance, Gemini Deep Think is poised for broader integration into platforms such as AI Studio and Google Cloud. Through API access, developers could use its reasoning capabilities for specialized applications, ranging from academic research to industrial problem-solving. This integration would enable organizations to harness the model's advanced capabilities in a more streamlined and accessible manner. The future of AI development, however, depends on addressing key trade-offs. Balancing intelligence, speed, and cost will be essential to making models like Gemini Deep Think scalable and efficient. Achieving this balance will determine the extent to which such technologies can be adopted across diverse industries, from education and healthcare to engineering and finance. By refining its computational efficiency and reducing processing times, Gemini Deep Think could unlock new possibilities for AI applications. Its ability to perform high-level reasoning tasks with remarkable accuracy positions it as a valuable tool for solving some of the most complex challenges in science, technology, and beyond. Gemini Deep Think represents a significant milestone in the evolution of artificial intelligence, showcasing its ability to perform advanced reasoning tasks with precision and accuracy. Its performance at the IMO underscores the potential of AI to rival human intelligence in domains that demand deep analytical thinking. However, the model's computational demands and extended processing times highlight areas that require improvement to ensure its practicality and scalability. As the field of AI continues to evolve, the focus will remain on optimizing efficiency, usability, and accessibility. By addressing these challenges, models like Gemini Deep Think could pave the way for fantastic advancements across a wide range of industries, shaping the future of artificial intelligence and its role in solving the world's most complex problems.

Gemini Deep Think : The Future of Precision AI Complex Problem-Solving?

What if the toughest problems humanity faces -- those that stump our brightest minds and stretch the limits of human ingenuity -- could be tackled by a single, purpose-built system? Enter Gemini Deep Think, an advanced artificial intelligence model designed to confront the most intricate challenges in mathematics, algorithm development, and design optimization. Unlike general-purpose AI tools, this system doesn't aim to be a jack-of-all-trades; instead, it's a master of depth and precision, engineered to excel where others falter. With features like parallel thinking and extended inference time, Gemini Deep Think pushes the boundaries of what AI can achieve, offering a glimpse into a future where even the most complex problems have solutions within reach. Prompt Engineering explores how Gemini Deep Think redefines AI-driven problem-solving, from its new innovations to its specialized applications in fields like advanced mathematics and software analysis. You'll discover how its unique capabilities -- such as iterative refinement and reinforcement learning -- enable it to outperform competitors and set new benchmarks for excellence. But this isn't just about technology; it's about the profound implications for industries and researchers tackling humanity's hardest questions. As we unpack its features and limitations, consider this: could a tool like Gemini Deep Think become the cornerstone of innovation in a world increasingly defined by complexity? Gemini Deep Think stands out due to its unique capabilities, which are specifically designed to address highly complex problems. These features enable it to operate at a level of sophistication that few AI systems can match: These features collectively make Gemini Deep Think a powerful tool for solving problems that demand advanced logical, mathematical, and creative reasoning. Gemini Deep Think has achieved remarkable results across a variety of benchmarks, solidifying its position as a leader in AI innovation. Its performance highlights include: These achievements underscore the model's potential to transform industries that rely on advanced reasoning, design, and optimization. Here are more guides from our previous articles and guides related to advanced AI models that you may find helpful. Gemini Deep Think is not a general-purpose AI but is specifically designed to excel in high-complexity, specialized scenarios. Its primary applications include: These specialized capabilities position Gemini Deep Think as an indispensable tool for researchers, engineers, and organizations operating at the forefront of innovation. While Gemini Deep Think excels in its specialized domains, it is important to recognize its limitations to ensure appropriate use: These constraints reflect the model's focus on delivering exceptional performance in targeted areas rather than attempting to address a broad range of applications. Gemini Deep Think integrates several new technologies that enhance its ability to solve complex problems: These innovations ensure that Gemini Deep Think remains at the cutting edge of AI technology, particularly for applications requiring rigorous reasoning and precision. When compared to other AI models, Gemini Deep Think consistently demonstrates superior performance in several key areas: These comparisons highlight the model's leadership in specialized problem-solving while also identifying opportunities for future enhancements. Gemini Deep Think represents a significant advancement in AI technology, offering unparalleled capabilities in solving complex problems across mathematics, coding, and design. Its specialized features, new innovations, and exceptional performance make it a valuable asset for researchers, engineers, and organizations tackling humanity's toughest challenges. While it is not designed for general-purpose use, its precision and depth ensure its relevance in addressing critical, high-stakes problems that demand advanced reasoning and creativity.

Gemini Deep Think AI Might Be at Critical Capability Levels Researchers Warn

Could a machine ever think so deeply that it rivals, or even surpasses, human ingenuity? With Google's Gemini Deep Think model, this question is no longer theoretical -- it's a pressing reality. Touted as a new leap in artificial intelligence, Gemini's capabilities extend far beyond solving complex equations or generating 3D models. It has cracked mathematical puzzles that stumped experts for decades and analyzed molecular structures with precision that could transform drug discovery. Yet, as researchers celebrate these achievements, they're also sounding the alarm: Gemini may have reached critical capability thresholds, where its potential for misuse is as staggering as its promise. The stakes have never been higher in the race to balance innovation with responsibility. This overview by Wes Roth provide more insights into the dual-edged nature of Gemini Deep Think, exploring its fantastic applications alongside the growing concerns it raises. How does a model capable of parallel thinking and reinforcement learning reshape fields like biology, cybersecurity, and engineering? And more importantly, what safeguards are needed to prevent it from becoming a tool for harm? By examining the intricate balance between progress and precaution, we uncover the profound implications of AI systems approaching the limits of their potential. As the lines between human and machine intelligence blur, the question isn't just what AI can do -- but whether we're ready for what comes next. The Gemini 2.5 model distinguishes itself through its ability to address problems previously deemed too intricate for AI systems. Its integration of parallel thinking and reinforcement learning allows it to process vast amounts of data and solve multifaceted challenges with exceptional efficiency. Some of its most new accomplishments include: These capabilities make Gemini Deep Think a fantastic tool for scientists, engineers, and researchers, allowing them to synthesize insights from extensive datasets and accelerate innovation across various disciplines. Despite its potential to transform research and development, access to Gemini Deep Think is tightly controlled. Users are limited to five interactions per day, a restriction aimed at managing the model's substantial computational demands and mitigating risks associated with its advanced functionalities. Furthermore, the model is exclusively available to premium subscribers, making sure that only a select group of users can use its capabilities. These limitations are not solely about resource allocation. They reflect broader concerns about the potential misuse of such a powerful tool. In fields like chemical and biological research, where technical expertise can be weaponized, restricting access is seen as a necessary safeguard to prevent unintended consequences. Take a look at other insightful guides from our broad collection that might capture your interest in Deep Thinking AI models. One of the most pressing concerns surrounding Gemini Deep Think is its ability to generate detailed technical knowledge in chemical, biological, radiological, and nuclear (CBRN) domains. This capability, while valuable for legitimate research, could be exploited by malicious actors to develop harmful technologies, such as bioweapons. The model's proficiency in synthesizing information from multiple research papers amplifies this risk, as it may inadvertently provide insights that could be misused. To mitigate these risks, experts are advocating for the implementation of: These measures are critical as AI systems like Gemini Deep Think approach what researchers describe as "critical capability thresholds," where their potential benefits are matched by equally significant risks. The rapid advancement of AI technologies has sparked widespread concern about their potential misuse. Beyond the risks in CBRN domains, there are growing fears about AI's applications in cybersecurity. Advanced models like Gemini Deep Think could be used to: These risks underscore the need for a balanced approach to AI development -- one that fosters innovation while prioritizing safety, ethical responsibility, and robust oversight. Despite the concerns, Gemini Deep Think has garnered praise for its practical applications across various fields. Early adopters have highlighted its ability to: These features make Gemini Deep Think an invaluable tool for professionals in fields ranging from engineering to scientific research. However, its benefits must be carefully weighed against the potential for misuse, emphasizing the importance of responsible development and deployment. As AI systems like Gemini Deep Think continue to evolve, the need for a cautious and deliberate approach becomes increasingly evident. While the model represents a significant milestone in artificial intelligence, it also highlights the ethical and safety challenges that accompany such advancements. By implementing proactive safeguards, conducting thorough risk assessments, and fostering a culture of responsibility, the AI community can ensure that these technologies are used to benefit society. Striking this balance is essential to harnessing the full potential of AI while minimizing its risks, paving the way for a future where innovation and responsibility coexist harmoniously.

Google Releases Two Versions of Gemini 2.5 Deep Think | PYMNTS.com

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. This model is a "significant improvement" over what Google announced in May at its Google I/O conference and is a variation of the one that achieved the gold-medal standard at this year's International Math Olympiad, the company said in a Friday blog post. The model released Friday to Google AI Ultra subscribers is designed for day-to-day use, reaching bronze-level performance on the 2025 IMO benchmark while operating faster, according to the post. Deep Think uses parallel thinking techniques, meaning it generates several ideas, considers them simultaneously and then arrives at the best answer, the post said. This model is likely to be used for iterative development and design, such as in web development; scientific and mathematical discovery; and algorithmic development and code, where it excels at tough coding problems, per the post. "Deep Think can help people tackle problems that require creativity, strategic planning and making improvements step-by-step...," the post said. Google also said Friday that it released the version of Google 2.5 Deep Think that achieved the gold-medal standard with a small group of mathematicians and academics. "We look forward to hearing how it could enhance their research and inquiry, and we'll use their feedback as we continue to improve this offering," the company said in the post. Deep Think mode was announced at Google I/O in May along with other Gemini AI models. The lineup included Gemini 2.5 Pro; Updated Gemini 2.5 Flash, described as a faster, more efficient model; Deep Think mode, described as a high-performance version of Gemini 2.5 Pro for complex reasoning; and Gemini SDK, which provides compatibility with Model Context Protocol (MCP) to let agents use a variety of tools.