Latest news with #2.5Pro
Yahoo
a day ago
- Business
- Yahoo
Google says its updated Gemini 2.5 Pro AI model is better at coding
Google on Thursday announced an update to its Gemini 2.5 Pro preview model that the company claims is better at certain programming tasks. The company's calling it an "updated preview," building on the upgrade to Gemini 2.5 Pro that Google announced around a month ago. Google says the model will roll out in general availability in a "couple of weeks," and is available starting today in its AI developer platforms AI Studio and Vertex AI and the Gemini app. "[Gemini 2.5 Pro] continues to excel at coding, leading on difficult coding benchmarks," Google writes in a blog post. "It also shows top-tier performance [on] highly challenging benchmarks that evaluate a model's math, science, knowledge, and reasoning capabilities." So what else is new? Google says it addressed feedback from its previous 2.5 Pro release, improving the model's style and structure. Now, 2.5 Pro can be "more creative with better-formatted responses," Google claims. This article originally appeared on TechCrunch at
Time of India
20-05-2025
- Business
- Time of India
Google I/O 2025: Gemini 2.5 Pro gets improved reasoning, audio features and multilingual support
At Google I/O 2025, the company announced new updates to its Gemini 2.5 model series adding more powerful reasoning, native audio output, security upgrades, and improved tools for developers. 'In March, we announced Gemini 2.5 Pro , our most intelligent model yet…Today, We're bringing new capabilities to 2.5 Pro and 2.5 Flash,' Google said, announcing the new updates. The upgraded Gemini 2.5 Pro model now tops performance charts, including WebDev Arena for coding and LMArena for human preference testing. It also features a 1 million-token context window, which allows it to handle longer inputs and video understanding tasks. Google said that thanks to LearnLM — a version of Gemini developed with educational experts — the model now leads in learning-related tasks as well. 'Educators and experts preferred Gemini 2.5 Pro over other models across a diverse range of scenarios,' the company said. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Ótimas notícias para os cidadãos brasileiros! Leia mais Undo Native audio, emotional dialogue and multilingual support Google also introduced native audio output for a more natural AI experience. Gemini can now speak with different tones, accents, and styles — such as a dramatic voice when telling a story. It can also: Detect user emotions and respond accordingly (Affective Dialogue) Ignore background noise (Proactive Audio) Handle more complex voice tasks (Thinking in the Live API) The text-to-speech tool now supports multiple speakers and over 24 languages, and it can switch between languages mid-conversation. These features will be available later today through the Gemini API. New 'Deep Think' for complex tasks Google said that it is testing an enhanced reasoning mode called Deep Think, which helps Gemini consider multiple answers before responding. It's aimed at tough challenges like advanced math and programming. 'We're starting to test an enhanced reasoning mode called Deep Think,' the company said. 'We're taking extra time to conduct more frontier safety evaluations and get further input from safety experts.' Deep Think is already leading benchmarks like the 2025 USAMO (math), LiveCodeBench (coding), and MMMU (multimodal reasoning). Gemini 2.5 Flash gets faster and more efficient Gemini 2.5 Flash, the lightweight version of the model, now uses 20–30% fewer tokens while improving performance across reasoning, code, and multimodal tasks, the company announced. It is now available in the Gemini app, Google AI Studio , and Vertex AI. A general release of the updated model is expected in early June, with 2.5 Pro following soon after. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Tom's Guide
20-05-2025
- Business
- Tom's Guide
Gemini just saw a huge upgrade to its AI model — here's everything new you can do
Google Gemini is stepping up its game. Announced at Google's yearly I/O event, the AI tool has just gotten one of its biggest updates ever, seeing improvements across multiple models and bringing in new features. This covers everything from improvements in coding and web design, to boosts in model efficiency and a brand new deep research feature. On top of this, Google has announced updates to its AI video generator with Veo 3, as well as announcing new AI plans and other improvements to its suite of AI tools. But, for now, let's focus on how Gemini looks different and all the new changes that are coming. A new feature announced at I/O for Gemini, Deep Think, is an enhanced reasoning mode. This uses new research techniques, enabling the model to consider multiple different hypotheses before responding. This is a concept that fits in well with reasoning models, where AI can think through a task with more detail. Google claims that 2.5 Pro Deep Think scored impressively on one of the hardest math benchmarks available, as well as leading on multiple AI testing systems for multimodal reasoning. Get instant access to breaking news, the hottest reviews, great deals and helpful tips. Deep Think won't be immediately available, and no release date has been announced yet. Google explained that they want to do further testing before they release this feature to the public. An update coming to Gemini in early June, Google announced improvements to its 2.5 Flash mode. This is the model designed for speed and low-cost tasks, build for simple prompting without the model needing to think too deeply. Google revealed that 2.5 Flash has been improved across key benchmarks for reasoning, multimodality, coding and long context. All of this while using 20 to 30% fewer tokens (computing power). The main update for 2.5 Pro from Google came early. Announced in the first week of May, this update improved Gemini 2.5 Pro's ability to build interactive web apps. This was a major improvement on the technology, and saw a big push for vibe coding (the ability to code through AI prompts). Announced at I/O, Google also claimed that the new 2.5 Pro model is now leading the popular coding leaderboard WebDev Arena, as well as leading multiple categories of the LMArena. These are tests of both how well the model can develop websites, and also their ability to take on tasks like image generation, and how efficient they are. Google claimed that with the improvements launched earlier this month, Gemini is now the leading model for learning, outperforming top models on every one of the five principles of learning science. One of the more interesting announcements out of the Gemini reveal is a feature where users can customise the dialogue of Gemini Live. This could be used to make Gemini more natural and expressive, allowing users to steer its tone, accent, and style of speaking. This will include a variety of new tools like affective dialogue, where the model detects emotion in your voice and replies accordingly. Proactive audio, where the model ignores background conversations and knows when to respond. And finally, deep thinking in Live conversations. This will first be made available in the Gemini API system for developers, but will likely then follow onto Gemini. Project Mariner will be coming to the Gemini API and Vertex AI. This is a research tool that enables human-agent interaction. In other words, it could allow Gemini to complete tasks across websites, like booking flights, completing forms, and following workflow summaries. For now, this will only be available to developers to experiment with and there is no detail of a future release on Gemini. Google claims that, with this latest update, Gemini 2.5 is the most secure AI model family they've made. This includes making improvements to protections against security threats and malicious instructions that could be embedded into the models. You'll likely not notice any changes here, but that just means it's working well! Further updates were announced specifically for developers using Gemini tools. These were broken down into three sections: Thought summaries, thinking budgets and MCP support. Thought Summaries are a new ability in Gemini API and Vertex AI, in which the model will summarize its raw thoughts and organize them with headers and key details. Thinking budgets allow developers to have more control over cost by balancing latency and quality, allowing them to control the number of tokens a model uses before it responds. Finally, MCP Support will make it easier to integrate the Gemini API with open-source tools. Google claims that it is also working on new approaches to improve the model and developer experience. This is why the tools will first be available to developers.
The Hindu
07-05-2025
- Business
- The Hindu
Gemini 2.5 Pro Preview unveiled ahead of Google I/O 2025
Google on Tuesday (May 6, 2025) unveiled its cutting-edge AI model, Gemini 2.5 Pro Preview. This is an upgraded version of its 2.5 Pro model, and it's coming out just before Google's developer event, Google I/O 2025. The 2.5 Pro Preview version builds on the strengths of its predecessor and makes improvements in code editing and developing complex agentic workflows. Developers can start using this updated Gemini 2.5 Pro in the Gemini API through Google AI Studio and Vertex AI. It's also available for users in the Gemini app, which powers features like Canvas. This means anyone can vibe code and build interactive web apps with just a single prompt. The 2.5 Pro is currently the top dog on the WebDev Arena Leaderboard, which measures how much people like a model's ability to create visually appealing and functional web apps.
India Today
04-05-2025
- Entertainment
- India Today
Gemini 2.5 Pro just won this popular 29 year old game, even Sundar Pichai is impressed
Google launched Gemini 2.5 Pro a month ago and claims that it is the "most intelligent AI model" to date. During the launch, the tech giant highlighted that this model is much better than its competition, including OpenAI o3 models, DeepSeek R1, Claude and more. While benchmarks (provided by Google) are living proof of it, a recent win against a 29-year-old video game, Pokmon Blue, also added another feather to its cap. Since these are just claims from Google, we wanted to see how good the model is, and here is what we found. But before you read our experience, the question is: why is winning against a video game a milestone for an AI model? Let's find out. advertisementGemini 2.5 Pro finishes Pokmon BlueFor context, Pokmon Blue (released in 1996) is known for its intricate gameplay mechanics, strategic combat, and open-world exploration—elements that pose significant challenges for AI systems. To perform well in the game, an AI must demonstrate abilities such as long-term planning, goal management, and visual navigation—core competencies in the pursuit of general artificial intelligence. Now that Gemini 2.5 Pro has won against the complexities of this game, the AI model has proved its title, "most intelligent model".Reacting to this win, CEO Sundar Pichai took to X (formerly Twitter), saying, "What a finish! Gemini 2.5 Pro just completed Pokmon Blue! advertisement To clarify, the Gemini Plays Pokmon livestream wasn't launched by Google itself, but by 'a 30-year-old software engineer unaffiliated with Google' who goes by the name Joel Z. Nonetheless, Google executives have shown enthusiastic support for the project. Logan Kilpatrick, product lead for Google AI Studio, shared an update last month, noting that Gemini was 'making great progress at completing Pokmon' and had 'earned its 5th badge (next best model only has 3 so far, though with a different agent harness).'During the launch, Google highlighted that one of the standout improvements in this model lies in its enhanced coding abilities, which have been described as 'a big leap over 2.0' with 'more improvements to come.' According to Google, '2.5 Pro excels at creating visually compelling web apps and agentic code applications, along with code transformation and editing.'In recognised industry benchmarks for agentic coding, Gemini 2.5 Pro delivered a strong performance—scoring 63.8 per cent on SWE-Bench Verified using a custom agent setup—highlighting its proficiency in complex software engineering tasks. Now that we are comparing, Anthropic's Claude AI model has also been in the race to beat another Pokmon version, Red. But it has not been successful so February, Anthropic showcased the strides its Claude AI models were making in Pokmon Red, noting that Claude's 'extended thinking and agent training' gave it 'a major boost' when tackling 'more unexpected' tasks, such as playing a classic video game. While Claude has made notable progress, it has yet to complete Pokmon as it may be, Gemini's performance doesn't yet signal true general intelligence. The developer still lends a hand from time to time—intervening to fix bugs or restrict certain actions, such as overusing escape items. He maintains that no direct walkthroughs or step-by-step guidance are provided, aside from a one-off case involving a known still an open question whether Gemini could manage the same feat entirely on its own. Nevertheless, its ability to navigate a game as intricate as Pokmon Blue—even with some support—demonstrates the remarkable potential of large language models when deployed within a carefully structured environment.
