Image Credit: mitchel3uo@Unspalsh

Google recently announced that it is expanding its Gemini 2.0 family to offer developers and users with advanced AI capabilities. With improved speed and reasoning, Gemini 2.0 Flash is now available through the Gemini API, Google AI Studio, and Vertex AI. Google also introduced Gemini 2.0 Pro Experimental, which is said to be their most advanced model for complicated prompts and code. It is currently available to Gemini Advanced users through the Gemini app and in Google AI Studio, with a context window of two million tokens. Additionally, the company announced the Gemini 2.0 Flash-Lite, which is currently in public preview and achieves a balance between high performance and cost effectiveness. Google highlighted its commitment to AI safety by implementing automated security evaluations and reinforcement learning techniques.

According to Reuters, Google has made a significant breakthrough in its AI products by incorporating its experimental Gemini 2.0 Flash Thinking model into the Gemini app. The model can now explain its responses to complex queries due to its addition of reasoning skills, which separates it from competitors like OpenAI and DeepSeek.
Additionally, Google is adding more cost-effective models to its Gemini AI lineup. In response to the positive feedback of version 1.5, the Gemini 2.0 Flash-Lite is made to be even affordable than the original Flash. The move aims to compete with cheaper alternatives such as DeepSeek.

A new version of the Gemini 2.0 Pro model is now in experimental testing, while Gemini 2.0 Flash, which initially debuted in December, is now accessible to everyone. As stated by Google DeepMind CTO Koray Kavukcuoglu, the new Pro model offers improved performance, especially in complex tasks and coding. DeepSeek’s last training run cost less than $6 million, which is considerably less than predictions for US AI companies, according to Reuters, which emphasizes increasing scrutiny around AI research costs. Investors are becoming concerned about Alphabet’s growing capital expenditures, as evidenced by the company’s shares plunging after a predicted 29% rise in spending.

The Verge reports that Google is doing significant advancements in its AI strategy, with plans to invest $75 billion this year, up from $32.3 billion in 2023, in AI as well as related technologies. The goal of this substantial investment is to maintain competitiveness in the rapidly evolving AI industry against competitors like Microsoft, Meta, and Amazon-backed Anthropic. Together with a more interactive version that can be integrated with programs like YouTube, Search, and Google Maps, the Gemini 2.0 Flash Thinking model is now available via the model selection in the Gemini app on desktop and mobile devices.

Users now have more device versatility thanks to this upgrade. In addition, Google introduced Gemini 2.0 Pro, an experimental version that offers better performance for math and coding jobs. Gemini 2.0 Pro is promoted as Google’s most powerful model to date and is available to Advanced Gemini users and also those who have access to AI Studio and Vertex AI. The low-cost, high-performance Gemini 2.0 Flash-Lite version was released to complete the updates. Flash-Lite, which offers better benchmark performance and similar speed and cost to 1.5 Flash. It is currently available for public on Google AI Studio and Vertex AI.

Resources: GoogleBlog, Reuters, TheVerge