Google unveils Gemini 2.0, its new AI model for the agentic era

Google this week unveiled Gemini 2.0, the latest version of its flagship artificial intelligence (AI) model, which it claimed is twice as fast as its earlier avatar.

Gemini 2.0 can generate images and audio across languages, and can assist in Google searches and coding projects, the company said pointing out that its “next era of models built for this new agentic era…will enable us to build new AI agents that bring us closer to our vision of a universal assistant.”

“If Gemini 1.0 was about organizing and understanding information, Gemini 2.0 is about making it much more useful,” chief executive officer (CEO) Sundar Pichai said,

Gemini’s new capabilities “make it possible to build agents that can think, remember, plan, and even take action on your behalf,” said Tulsee Doshi, a director of product management at the company, said in a media briefing.

In a blog post, Google said Gemini 2.0 “can handle images, video, and audio inputs while generating outputs that mix text with natively created images and steerable text-to-speech audio.”

Unlike its competitors such as OpenAI’s GPT models, Gemini 2.0 allows developers to access tools like Google Search and execute code natively, it added.

Meanwhile, Google also released the Gemini 2.0 Flash tool, touted to be an improvement on its predecessor, Gemini 1.5 Flash.

One of the key advancements in Gemini 2.0 Flash is its ability to integrate seamlessly with existing applications and systems, providing developers with robust tools to create more dynamic and responsive AI-driven applications.

This includes better handling of natural language understanding and generation, more sophisticated image processing, and new functionalities in speech synthesis and recognition

The new model has improved performance across coding, math, reasoning, and factual accuracy benchmarks.

It is available as an experimental feature through platforms like Google AI Studio and Vertex AI, with general availability planned for January.

“Starting today, our Gemini 2.0 Flash experimental model will be available to all Gemini users. We’re also launching a new feature called Deep Research, which uses advanced reasoning and long context capabilities to act as a research assistant, exploring complex topics and compiling reports on your behalf. It’s available in Gemini Advanced today,” Pichai said.

Developers will also have access to a Multimodal Live API tool, which supports real-time audio and video streaming inputs.

The new features will be rolled out across Google products, including Android Studio, Chrome DevTools, and Firebase, starting in early 2025.

The release of Gemini 2.0 comes a day after the Information reported that Google flagged Microsoft’s exclusive partnership with OpenAI to the US Federal Trade Commission, citing potential anti-competitive practices in the AI and cloud markets.

According to the agreement, Microsoft Azure is the sole cloud service provider for OpenAI’s technologies, such as GPT and DALL-E.

The report said Google has requested the FTC to intervene in the deal and stated that this partnership limited access to OpenAI’s advanced AI models for businesses using other cloud platforms.

Companies like Google and Amazon, which also offer cloud services, have expressed dissatisfaction with the inability to host OpenAI’s models independently.

They maintain that those relying on OpenAI’s technologies are forced to use Microsoft servers, limiting flexibility and increasing costs for those operating on different cloud platforms.