Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More<\/em><\/p>\n\n\n\n Google is moving closer to its goal of a \u201cuniversal AI assistant\u201d that can understand context, plan and take action.\u00a0<\/p>\n\n\n\n Today at Google I\/O, the tech giant announced enhancements to its Gemini 2.5 Flash \u2014 it\u2019s now better across nearly every dimension, including benchmarks for reasoning, code and long context \u2014 and 2.5 Pro, including an experimental enhanced reasoning mode, \u2018Deep Think,\u2019 that allows Pro to consider multiple hypotheses before responding.\u00a0<\/p>\n\n\n\n \u201cThis is our ultimate goal for the Gemini app: An AI that\u2019s personal, proactive and powerful,\u201d Demis Hassabis, CEO of Google DeepMind, said in a press pre-brief.\u00a0<\/p>\n\n\n\n Google announced Gemini 2.5 Pro \u2014 what it considers its most intelligent model yet, with a one-million-token context window \u2014 in March, and released its \u201cI\/O\u201d coding edition earlier this month (with Hassabis calling it \u201cthe best coding model we\u2019ve ever built!\u201d).\u00a0<\/p>\n\n\n\n \u201cWe\u2019ve been really impressed by what people have created, from turning sketches into interactive apps to simulating entire cities,\u201d said Hassabis.\u00a0<\/p>\n\n\n\n He noted that, based on Google\u2019s experience with\u00a0AlphaGo, AI model responses improve when they\u2019re given <\/span>more time to think. This led DeepMind scientists to develop Deep Think, which uses Google\u2019s latest cutting-edge research in thinking and reasoning, including parallel techniques.<\/p>\n\n\n\n Deep Think has shown impressive scores on the hardest math and coding benchmarks, including the 2025 USA Mathematical Olympiad (USAMO). It also leads on LiveCodeBench, a difficult benchmark for competition-level coding, and scores 84.0% on MMMU, which tests multimodal understanding and reasoning.<\/p>\n\n\n\n Hassabis added, \u201cWe\u2019re taking a bit of extra time to conduct more frontier safety evaluations and get further input from safety experts.\u201d (Meaning: As for now, it is available to trusted testers via the API for feedback before the capability is made widely available.)<\/p>\n\n\n\n Overall, the new 2.5 Pro leads popular coding leaderboard WebDev Arena, with an ELO score \u2014 which measures the relative skill level of players in two-player games like chess \u2014 of 1420 (intermediate to proficient). It also leads across all categories of the LMArena leaderboard, which evaluates AI based on human preference.\u00a0<\/p>\n\n\n\n Since its launch, \u201cwe\u2019ve been really impressed by what [users have] created, from turning sketches into interactive apps to simulating entire cities,\u201d said Hassabis.\u00a0<\/p>\n\n\n\n Also today, Google announced an enhanced 2.5 Flash, considered its workhorse model designed for speed, efficiency and low cost. 2.5 Flash has been improved across the board in benchmarks for reasoning, multimodality, code and long context \u2014 Hassabis noted that it\u2019s \u201csecond only\u201d to 2.5 Pro on the LMArena leaderboard. The model is also more efficient, using 20 to 30% fewer tokens.<\/p>\n\n\n\n Google is making final adjustments to 2.5 Flash based on developer feedback; it is now available for preview in Google AI Studio, Vertex AI and in the Gemini app. It will be generally available for production in early June.<\/p>\n\n\n\n Google is bringing additional capabilities to both Gemini 2.5 Pro and 2.5 Flash, including native audio output to create more natural conversational experiences, text-to-speech to support multiple speakers, thought summaries and thinking budgets.\u00a0<\/p>\n\n\n\n With native audio input (in preview), users can steer Gemini\u2019s tone, accent and style of speaking (think: directing the model to be melodramatic or maudlin when telling a story). Like Project Mariner, the model is also equipped with tool use, allowing it to search on users\u2019 behalf.\u00a0<\/p>\n\n\n\n Other experimental early voice features include affective dialogue, which gives the model the ability to detect emotion in user voice and respond appropriately; proactive audio that allows it to tune out background conversations; and thinking in the Live API to support more complex tasks.\u00a0<\/p>\n\n\n\n New multiple-speaker features in both Pro and Flash support more than 24 languages, and the models can quickly switch from one dialect to another. \u201cText-to-speech is expressive and can capture subtle nuances, such as whispers,\u201d Koray Kavukcuoglu, CTO of Google DeepMind, and Tulsee Doshi, senior director for product management at Google DeepMind, wrote in a blog posted today.\u00a0<\/p>\n\n\n\n Further, 2.5 Pro and Flash now include thought summaries in the Gemini API and Vertex AI. These \u201ctake the model\u2019s raw thoughts and organize them into a clear format with headers, key details, and information about model actions, like when they use tools,\u201d Kavukcuoglu and Doshi explain. The goal is to provide a more structured, streamlined format for the model\u2019s thinking process and give users interactions with Gemini that are simpler to understand and debug.\u00a0<\/p>\n\n\n\n Like 2.5 Flash, Pro is also now equipped with \u2018thinking budgets,\u2019 which gives developers the ability to control the number of tokens a model uses to think before it responds, or, if they prefer, turn its thinking capabilities off altogether. This capability will be generally available in coming weeks.<\/p>\n\n\n\n Finally, Google has added native SDK support for Model Context Protocol (MCP) definitions in the Gemini API so that models can more easily integrate with open-source tools.<\/p>\n\n\n\n As Hassabis put it: \u201cWe\u2019re living through a remarkable moment in history where AI is making possible an amazing new future. It\u2019s been relentless progress.\u201d<\/p>\n Daily insights on business use cases with VB Daily<\/strong><\/p>\n If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI.<\/p>\n \n\t\t\t\t\tThanks for subscribing. Check out more VB newsletters here.\n\t\t\t\t<\/p>\n An error occured.<\/p>\n<\/p><\/div>\n
\n<\/div>\u2018Deep Think\u2019 scores impressively on top benchmarks<\/h2>\n\n\n\n
Important updates to Gemini 2.5 Pro, Flash<\/h2>\n\n\n\n
![\"\"\/](\"https:\/\/venturebeat.com\/wp-content\/themes\/vb-news\/brand\/img\/vb-daily-phone.png\")
\n\t\t\t\t<\/div>\n<\/p><\/div>\n<\/div>\t\t\t<\/div>\r\n