Google releases Gemini 2.5 Deep Think for AI Ultra subscribers
How did your country report this? Share your view in the comments.
Diverging Reports Breakdown
Google releases Gemini 2.5 Deep Think for AI Ultra subscribers
Gemini 2.5 Deep Think is Google’s most powerful Gemini model. It is available only to users of Google’s $250 AI Ultra plan. It uses more compute resources than other models and is designed for the most complex queries. Google says it is more adept at design aesthetics, scientific reasoning, and coding.
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.
Credit: Google
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.
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.
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 launches Gemini 2.5 Deep Think model, beats OpenAI o3 and Grok 4 in performance
Google announced the rollout of the Gemini 2.5 Deep Think model in the Gemini app for Google AI Ultra subscribers. This new model beats OpenAI’s o3 and xAI’s Gork 4 models in several key benchmarks. Google is also planning to release Deep Think with and without tools to a set of trusted developers via the Gemini API in the coming weeks. Google noted that Deep Think automatically works with tools such as code execution and Google Search.
Today, Google announced the rollout of the Gemini 2.5 Deep Think model in the Gemini app for Google AI Ultra subscribers. This new model beats OpenAI’s o3 and xAI’s Gork 4 models in several key benchmarks.
Back in May, at Google I/O 2025, Google first announced the Gemini 2.5 Pro Deep Think mode, which used new research techniques to consider multiple hypotheses before responding. Today’s rollout is an updated Deep Think model that includes improvements based on feedback from early trusted testers and recent research breakthroughs. Google claims that this updated Deep Think model is a significant improvement over the version first announced at I/O.
Google also revealed that this new Deep Think model is a variation of the model that recently achieved the gold-medal standard at this year’s International Mathematical Olympiad (IMO). To make it suitable for everyday usage, Google has made it faster, and it, therefore, reaches only Bronze-level performance on the 2025 IMO benchmark.
As seen in the benchmarks in the screenshot above, Gemini 2.5 Deep Think is shown to achieve excellent performance across LiveCodeBench V6, Humanity’s Last Exam, IMO 2025, and AIME 2025.
Google AI Ultra subscribers can now use Deep Think in the Gemini app with a fixed number of prompts per day by using the ‘Deep Think’ option that appears in the prompt bar when the Gemini 2.5 Pro model is selected. Google noted that Deep Think automatically works with tools such as code execution and Google Search. Google is also planning to release Deep Think with and without tools to a set of trusted developers via the Gemini API in the coming weeks.
The big question now is whether the upcoming OpenAI GPT-5 model will outperform Gemini 2.5 Deep Think in key benchmark results.
Source: https://arstechnica.com/ai/2025/08/google-releases-gemini-2-5-deep-think-for-ai-ultra-subscribers/