Is Google's AI model a 'bargain' for users?

Leading in important rankings, Gemini 2.5 Pro has many outstanding improvements compared to the old version, as well as other AI chatbots on the market today.

This past weekend, Google unexpectedly made a beta version of Gemini 2.5 Pro, the company's newest and smartest model, available to all Gemini users. Previously, this model was only available to Gemini Advanced subscribers, priced at $20 per month.

Gemini 2.5 Pro integrates thinking capabilities rather than providing them separately. Although it's a beta version, this model already supports several features such as applications/extensions, file uploads, and Canvas.

In the prestigious AI ranking by LMArena, Gemini 2.5 Pro is leading. This model also achieved the best results in math (AIME 2025) and science (GPQA diamond) tests without the need for expensive testing techniques.

A free trial of Gemini 2.5 Pro is available to all users. Photo: Google.

Breakthroughs in programming

Google considers Gemini 2.5 Pro a "leap forward" compared to version 2.0. This model achieved a score of 18.8%, the highest among models that did not use external tools in the Humanity's Last Exam, a dataset designed by hundreds of experts to assess the limits of human knowledge and thinking.

This version also excels in creating intuitive web applications and automated coding, achieving 63.8% in the SWE-Bench automated programming test, significantly higher than Gemini 2.0 and prominent competitors such as OpenAI o3-mini and Grok 3 Beta.

Comparison of Gemini 2.5 Pro's performance metrics with the smartest chatbots currently available. Image: Google.

Additionally, Gemini 2.5 Pro's context window, already large at 1 million tokens (equivalent to 700 document pages), is about to double in size. Second place goes to Claude with 200,000 tokens.

This means the model can understand massive datasets and handle complex problems from multiple sources of information, including text, audio, images, video , and even source code repositories.

According to Theo Browne, YouTuber and CEO of Ping Labs, with such a large context window, the model's processor can contain almost the entire source code of a small software project.

Content creators challenged Gemini 2.5 Pro with various demanding tasks such as creating websites and games using only a single command.

Multi-application integration

Beyond supporting programmers, Gemini 2.5 Pro will become a powerful tool for content creators and designers. The model performs better with video, audio, and image creation tasks, and also integrates Canvas, a feature that helps AI draw diagrams, design UI, and suggest concept ideas.

Perhaps the most widespread application of this model is in the field of education, with its large storage capacity and ability to process numerous documents without interruption. Students can personalize their learning paths and create engaging images and videos for their presentations.

Currently, Gemini 2.5 Pro is supported directly on gemini.google.com and Google AI Studio. In the future, the model will be integrated into Vertex AI, Google Cloud's premium platform, enabling businesses to deploy and manage AI models on a large scale.

In Vietnam, users are debating Gemini's programming capabilities. Doan Kien, founder of Seodo Media, claims that the Gemini 2.5 Pro is currently in the leading position. However, his post has received many opposing comments, arguing that Claude is the one that deserves that position.

Comparing the capabilities of AI chatbots. Photo: Peter Yang/X

In the near future, Google will focus on improving the programming capabilities of the Gemini 2.5 Pro. Although there is no end date for the trial period or pricing policy announcement, many users predict that when it is officially released, the Gemini 2.5 Pro will be more expensive than other models, but will come with more flexible options.

Source: https://znews.vn/mo-hinh-ai-cua-google-co-phai-mon-hoi-cho-nguoi-dung-post1542121.html