Google announced the launch of Gemini 2.0, the latest AI model designed to transform the way people interact with and benefit from artificial intelligence.
Drawing on the revolutionary capabilities of Gemini 1.0This next-generation model introduces advancements in multimodality, agentic capabilities, and tool integration, paving the way for a new era of AI-driven solutions for individuals, businesses, and developers.
“For more than 26 years, Google has worked to organize the world’s information and make it universally accessible and useful” said Sundar Pichai, CEO of Google and Alphabet.
“Gemini 2.0 takes this mission further by enabling AI to not only understand and organize information, but also act intelligently on it, bringing us closer to the vision of a profoundly useful universal assistant in daily life.”
The Gemini 2.0 difference
Gemini 2.0 builds on the success of its predecessor as the first natively multimodal model, which integrates and processes information in the form of text, images, audio, video and code.
Now, with Gemini 2.0 Flash — its first experimental model — the technology allows:
- Multimodal inputs and outputs: Gemini 2.0 supports natively generated multilingual images, text, and audio, creating a richer, more interactive experience.
- Integration of native tools: The template can seamlessly access tools like Google Search, code execution platforms, and even third-party user-defined functions.
- Improved performance: Gemini 2.0 Flash outperforms its predecessor (1.5 Pro) on key criteria while offering twice the speed.
Developers can now access the experimental model via the Gemini API in Google AI Studio and Vertex AI, with full general availability planned for January 2025.
Transforming AI Experiences
Gemini 2.0 brings more than incremental upgrades: it’s a step toward “agentic” AI, a concept in which models can anticipate multiple steps, execute tasks, and collaborate intelligently with users.
This capacity underpins projects such as:
- Project Astra: a prototype exploring future universal assistant capabilities.
- Project Mariner: a revolutionary initiative focused on redefining human-agent interactions.
- Jules: an AI-powered code assistant designed for developers.
“Gemini 2.0 is not only smarter, it’s action-oriented. » said Demis Hassabis, CEO of Google DeepMind. “It is designed to understand your needs, think strategically and help you proactively, opening the door to entirely new possibilities for AI-based solutions.” »
AI advances in the Google ecosystem
Gemini 2.0’s capabilities are already reshaping Google’s core products, including search, which now leverages the technology’s advanced reasoning to handle complex queries and multimodal questions.
AI Overviews, one of the fastest growing search features, will soon integrate Gemini 2.0 to improve its ability to solve advanced math, multimodal query, and coding problems.
Gemini Assistant, available worldwide in chat-optimized Flash 2.0 versions, will bring even more dynamic functionality to users across desktop and mobile platforms. Further integration into the Google product suite is planned for early 2024.
Built on Google’s Full-Stack AI Leadership
Gemini 2.0 represents the culmination of Google’s ten years of investment in AI innovation.
Fully powered by Google’s custom Trillium TPUs, the model exemplifies efficiency, scalability, and performance. Trillium is now also available to customers, enabling developers to build next-generation applications at scale.
“Gemini 2.0 is more than a model: it is a step towards the future of AI. » Pichai concluded. “By combining multimodal capabilities with agentic intelligence, it paves the way for a universal AI assistant that empowers everyone to do more.” We can’t wait to see what developers and users create with it. »
View the full announcement here.
Don’t miss the important articles of the week. Subscribe to techbuild weekly summary for updates
Google announced the launch of Gemini 2.0, the latest AI model designed to transform the way people interact with and benefit from artificial intelligence.
Drawing on the revolutionary capabilities of Gemini 1.0This next-generation model introduces advancements in multimodality, agentic capabilities, and tool integration, paving the way for a new era of AI-driven solutions for individuals, businesses, and developers.
“For more than 26 years, Google has worked to organize the world’s information and make it universally accessible and useful” said Sundar Pichai, CEO of Google and Alphabet.
“Gemini 2.0 takes this mission further by enabling AI to not only understand and organize information, but also act intelligently on it, bringing us closer to the vision of a profoundly useful universal assistant in daily life.”
The Gemini 2.0 difference
Gemini 2.0 builds on the success of its predecessor as the first natively multimodal model, which integrates and processes information in the form of text, images, audio, video and code.
Now, with Gemini 2.0 Flash — its first experimental model — the technology allows:
- Multimodal inputs and outputs: Gemini 2.0 supports natively generated multilingual images, text, and audio, creating a richer, more interactive experience.
- Integration of native tools: The template can seamlessly access tools like Google Search, code execution platforms, and even third-party user-defined functions.
- Improved performance: Gemini 2.0 Flash outperforms its predecessor (1.5 Pro) on key criteria while offering twice the speed.
Developers can now access the experimental model via the Gemini API in Google AI Studio and Vertex AI, with full general availability planned for January 2025.
Transforming AI Experiences
Gemini 2.0 brings more than incremental upgrades: it’s a step toward “agentic” AI, a concept in which models can anticipate multiple steps, execute tasks, and collaborate intelligently with users.
This capacity underpins projects such as:
- Project Astra: a prototype exploring future universal assistant capabilities.
- Project Mariner: a revolutionary initiative focused on redefining human-agent interactions.
- Jules: an AI-powered code assistant designed for developers.
“Gemini 2.0 is not only smarter, it’s action-oriented. » said Demis Hassabis, CEO of Google DeepMind. “It is designed to understand your needs, think strategically and help you proactively, opening the door to entirely new possibilities for AI-based solutions.” »
AI advances in the Google ecosystem
Gemini 2.0’s capabilities are already reshaping Google’s core products, including search, which now leverages the technology’s advanced reasoning to handle complex queries and multimodal questions.
AI Overviews, one of the fastest growing search features, will soon integrate Gemini 2.0 to improve its ability to solve advanced math, multimodal query, and coding problems.
Gemini Assistant, available worldwide in chat-optimized Flash 2.0 versions, will bring even more dynamic functionality to users across desktop and mobile platforms. Further integration into the Google product suite is planned for early 2024.
Built on Google’s Full-Stack AI Leadership
Gemini 2.0 represents the culmination of Google’s decade-long investment in AI innovation.
Fully powered by Google’s custom Trillium TPUs, the model exemplifies efficiency, scalability, and performance. Trillium is now also available to customers, enabling developers to create next-generation applications at scale.
“Gemini 2.0 is more than a model: it is a step towards the future of AI. » Pichai concluded. “By combining multimodal capabilities with agentic intelligence, it paves the way for a universal AI assistant that empowers everyone to do more.” We can’t wait to see what developers and users create with it. »
View the full announcement here.
Don’t miss the important articles of the week. Subscribe to techbuild weekly summary for updates
