Google releases Gemini 2.0 models: 'We kicked off the agentic era'
In response to intense competition from DeepSeek, Google has unveiled its latest AI model, Gemini 2.0. The launch was officially announced on Google's Blog, detailing the enhanced features of Gemini 2.0 models. These new models boast improvements in efficiency, coding performance, AI safety, and cost-effectiveness.
New Offerings
The Gemini 2.0 series introduces several new offerings. Among them are:
- Gemini 2.0 Flash: A high-speed multi-modal model capable of processing voice, text, image, and video inputs, designed for large-scale tasks.
- Gemini 2.0 Pro (Experimental): Google's most advanced coding and complex reasoning model.
- Gemini 2.0 Flash-Lite: An affordable AI model that builds upon the capabilities of the previous Gemini 1.5 Flash.
The decision to release these models was driven by the growing competition from Chinese company DeepSeek, which disrupted the market with its DeepSeek R1 model. Google's response is part of a trend where various companies, including OpenAI, have launched new models to counter the competition.
Competition Response
OpenAI, in particular, swiftly released their latest model o3-mini at no cost. Additionally, they introduced autonomous AI agents like 'Operator' and 'Deep Research' in rapid succession. 
The Gemini 2.0 series has been equipped with various benchmarks to showcase advancements in high-frequency tasks, coding, complex reasoning, and long-context understanding. Google highlighted this release as a significant step in their AI evolution, stating:
"Our best model yet for coding performance and complex prompts."
"We kicked off the agentic era"
Availability and Access
While benchmarks provide insights into the technical capabilities of the models, the true test lies in their practical application by users. Gemini 2.0 Flash is accessible to Gemini Advanced users through the Gemini API on Google AI Studio, Vertex AI, and the Gemini app. Both Gemini 2.0 Pro (Experimental) and Gemini Flash-Lite can be obtained through Google AI Studio and Vertex AI.
