Google’s Upcoming AI ‘Gemini’ Has Launched In the Market With Its 3 Variants
By Last night Google launched its very own generative-AI, Gemini with three different size variations with each serving different specific purposes. This new language model was designed to understand, analyze, summarize and generate different types of contents. Google claims that this new language model at present is one of their most capable AI models.
Gemini is not only a general AI model but a family of AI models consisting of three different sizes. Starting with the smallest size LLM model Gemini Nano, then Gemini Pro and lastly, Gemini Ultra. All the three variations of this AI model have serves specific purpose.
As per the company, the Gemini model launched in the market is the first version of this AL model, version 1.0. This LLM model offers flexibility to run information from smartphones to data centers.
Google CEO, Sundar Pichai explain in his official blog post, “We designed Gemini to be natively multimodal, pre-trained from the start on different modalities. Then we fine-tuned it with additional multimodal data to further refine its effectiveness.”
Gemini AI model’s Nano version will be available on Pixel 8 Pro with enhanced features. These features included are Smart Reply on Gboard initially this feature was implemented in Whatsapp and Summarizing the Recorder app.
Gemini AI model series will also expand with several Google products as well as services including Duet AI, Chrome, Search and Ads. Gemini AI model is already being incorporated with Google Search that will speed up the SGE or Search Generative Experience. This inclusion has resulted in “a 40 per cent reduction in latency in English in the US, alongside improvements in quality.”
Gemini Ultra is one of the most powerful LLM that can excel at handling complex tasks and best suited for enterprise applications and data centers. Google describe this LLM model as exceeding “current state-of-the-art results on 30 of the 32 widely-used academic benchmarks used in large language model (LLM) research and development.”
This means this model exceeds expectations in performing standard tests and can tackle complex tasks. This model is also set to release after completing its testing phase.
Gemini Pro another variant of the LLM model Gemini designed specifically to power AI chatbot, Bard. This advanced technology will deliver fast responses and understand complex queries. This model is also the driving force behind some of Google AI services including Bard.Gemini Nano is the lightweight version of Gemini AI that was designed for smartphones especially Google’s Pixel 8 model. This LLM model efficiently performs on-device tasks even offline. From chat reply suggestion to summarizing text content this device offers smooth responsive experience for its users.
