Gemini 2.0 From Google Is Efficiency
On Wednesday, Google officially announced Gemini 2.0 , the company’s latest update to its flagship artificial intelligence model. In particular, Google is releasing an “experimental version of Gemini 2.0 Flash,” which replaces Gemini 1.5 Flash. Google’s Flash models are “lightweight” models, designed for tasks that don’t require the most powerful AI models and are more focused on efficiency. However, Google says that Gemini 2.0 Flash not only improves Flash models like Gemini 1.5 Flash, but also more powerful models like Gemini 1.5 Pro .
How is Gemini 2.0 Flash different from other models?
Google says 2.0 Flash outperforms 1.5 Flash and 1.5 Pro in a number of categories, including the MMLU-Pro general test, three different coding tests, fact checking, two math tests, reasoning, two image tests, and video tests. Some of these wins were close to 1.5 Pro’s performance, but others showed significant improvements, such as a 7.5-point increase in the Natural2Code test or a nine-point increase in the HiddenMath test. 1.5 Pro still beats 2.0 Flash in audio testing (40.1% vs. 39.2%) and long context testing (82.6% vs. 69.2%).
In addition to these improvements, Google says Flash 2.0 supports new multimodal output, such as AI-generated images combined with text and text-to-speech. Additionally, it can use Google search, run code, and other third-party functions.
Where will you see Gemini 2.0 Flash?
You’ll probably see Gemini 2.0 Flash a lot – whether you know it or not. The company announced that it will use Gemini 2.0 for search, specifically for AI reviews . Google’s initial introduction of AI search summaries was clearly a complete failure . However, the company is optimistic about the feature: Google says Gemini 2.0 will allow AI reviews to handle more complex topics and multi-step queries, as well as new features such as complex math calculations, multimodal questions (i.e. queries from text, images, documents). etc.), as well as coding.
Flash 2.0 will also be coming to the Gemini app. In fact, it is already available on desktop and mobile web. You will simply need to select a model from the drop-down menu before testing it.
Google wants AI to do all the work for you
Google is touting Flash 2.0 as part of its “Age of Agents.” This means Google wants its products to do more on your behalf, whether that’s analyzing an issue or your environment or actually performing a task for you. The company says it is working on updates to Project Astra, Google’s research arm responsible for developing a “universal AI assistant”; Project Mariner, a Chrome extension that uses artificial intelligence to help you surf the web; and Jules, an artificial intelligence company agent who helps developers write code.
Google has also highlighted a new feature called Deep Research , an AI-powered research assistant that aims to analyze things and create reports for you. You give the bot a topic or question you want to study, and it develops a research plan that you can approve or revise. Once approved, it scans the Internet for sources and produces a full report that can be exported to Google Docs. Like AI Reviews, it includes links to the sources it’s taken from so you can check them out for yourself.