Google is reportedly gearing up to launch its next frontier AI model in December. Purportedly called Gemini 2.0, this large language model – if launched – will come a year after the tech giant announced Gemini 1.0.
According to a report by The Verge, Google wants to “widely release” Gemini 2.0 in December, however, the model is said to not showing the prowess it is expected to demonstrate. The “model isn’t showing the performance gains the Demis Hassabis-led team had hoped for,” the report said.
The report, however, added that it appears that this “trend [of low perfomrance gains] is happening across companies developing leading, large models.”
Google Gemini AI models
Google’s Gemini AI models have seen a rapid evolution since the launch of the first model. Here’s a breakdown of the key releases:
December 2023: Google launched Gemini 1.0 and Gemini 1.0 Pro, marking Google’s entry into the large language model space. Gemini Pro was made available through Bard (before its rebranding to Gemini) and a Nano version was integrated into the Pixel 8 Pro.
February 2024: The company launched Gemini 1.5 with a significantly expanded context window. Google said that the model enhances its ability to handle and understand larger amounts of information.
May 2024: At Google I/O, the company announced Gemini 1.5 Flash, along with Gemma 2. These models were positioned as more accessible and efficient options for developers. Gemma 2 was officially released a month later.
September 2024: Updated versions of Gemini 1.5 Pro and Gemini 1.5 Flash were released, featuring improvements in quality and capabilities.
Project Astra: At I/O 2024, Google showcased Project Astra, a multimodal AI assistant capable of real-time interaction through text, audio and video.