Google is launching a free Gemini 2.0 Flash Thinking model, putting pressure on its premium OpenAI strategy

Google It has quietly released a major update to its popular AI model, Gemini, which now explains its inference process, sets new performance records on mathematical and scientific tasks, and offers a free alternative to premium OpenAI services.

New Gemini 2.0 Flash Thinking Modelreleased on Tuesday in Google Artificial Intelligence Studio Under the name “experimental”Exp-01-21“, and obtained a percentage of 73.3% American Invitational Mathematics Examination (AIME) and 74.2% on GPQA Diamond Science standard. These results show clear improvements over previous AI models and demonstrate Google’s increasing power in forward-thinking.

“We have been pioneering these types of planning systems for more than a decade, starting with programs like AlphaGo, and it is exciting to see the powerful combination of these ideas with more capable underlying models,” he wrote. Demis HassabisCEO of Google DeepMind, in a post on X.com (formerly Twitter).

Our latest update to the Gemini 2.0 Flash Thinking model (available here: https://t.co/Rr9DvqbUdO) He achieved 73.3% in the AIME (Mathematics) and 74.2% in the GPQA Diamond Standards (Science). Thanks for all your feedback, this represents very rapid progress since our first release just last… pic.twitter.com/cM1gNwBoTO

– Demis Hassabis (@demishassabis) January 21, 2025

Gemini 2.0 Flash Thinking breaks records by processing 1 million tokens

The most striking feature of this model is its ability to process up to one million text symbols – five times more than OpenAI’s o1 Pro model – While maintaining faster response times. This expanded context window allows the model to analyze multiple research papers or large-scale datasets simultaneously, a capability that could change the way researchers and analysts work with large amounts of information.

“As a first experiment, I took several religious and philosophical texts and asked Gemini 2.0 Flash Thinking to weave them together, extracting new and unique insights.” Dan MacOne AI researcher who tested the model in a… Published on X.com. “970,000 codes were processed in total. The output is pretty incredible.

The release comes at a critical moment in the development of the AI ​​industry. OpenAI recently announced o3 modelwhich achieved a score of 87.7% on the GPQA Diamond standard. However, Google’s decision to offer its model for free during beta testing (with usage limits) could attract developers and organizations looking for alternatives to OpenAI monthly subscription is $200.

Google offers Gemini 2.0 Flash Thinking for free with built-in code implementation

Jeff DeanGoogle DeepMind’s chief scientist emphasized the improvements in the model’s reliability: “We are continuing to iterate, with higher reliability and fewer discrepancies between the model’s ideas and the final answers.” books.

The model also includes native code execution capabilities, allowing developers to run and test code directly within the system. This feature, combined with improved inconsistency protections, positions Gemini 2.0 Flash Thinking as a serious contender for both research and commercial applications.

Industry analysts point out that Google’s focus on explaining its inference process could help address growing concerns about the transparency and reliability of artificial intelligence. Unlike traditional “black box” models, Gemini 2.0 Flash Thinking displays its working, making it easier for users to understand and verify its conclusions.

We continue to iterate, with higher reliability and fewer discrepancies between model ideas and final answers.

Check it as Gemini-2.0-flash-thinking-exp-01-21 on https://t.co/sw0jY6k74m

– Jeff Dean (@JeffDean) January 21, 2025

AI transparency becomes the new battleground as Google challenges OpenAI

The model has already taken first place in Leaderboard in Chatbot Arenaa preeminent standard for AI performance, is a leader in categories including challenging prompts, programming, and creative writing.

However, questions remain about the model’s real-world performance and its limitations. Although benchmark results provide valuable metrics, they do not always translate directly into practical applications. Google’s challenge is to convince enterprise customers that its free offerings can match or exceed the capabilities of premium alternatives.

As the AI ​​arms race intensifies, Google’s latest release signals a shift in strategy: combining advanced capabilities with accessibility. It remains to be seen whether this approach will help bridge the gap with OpenAI, but it certainly gives technical decision-makers a compelling reason to reconsider their AI partnerships.

For now, one thing is clear: the era of artificial intelligence that can show its work, available to anyone with a Google account, has arrived.

Daily insights into business use cases with VB Daily

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 organizational transformations to hands-on deployments, so you can share insights to maximize ROI.

Read our privacy policy

Thanks for subscribing. Check more VB newsletters here.

An error occurred.