After much anticipation, Gemini Ultra, the most powerful version of Google's AI model is finally available to the public. In addition to the release, Google rebranded a few of its AI products to "Gemini."
Here’s the full breakdown of the release and the changes.
It’s a lot of things.
First, Gemini is the name of Google’s foundational AI model (like OpenAI's GPT or Meta's Llama 2) and comes in three versions:
Now, here’s where it gets a bit tricky.
Gemini is also the name of Google’s free subscription service for access to the Gemini Pro 1.0 model. For access to Ultra, you’ll have to upgrade to Gemini Advanced.
Lastly, both Bard – Google’s generative AI tool on Google Workspace – and Duet AI from Google Cloud have been rebranded as "Gemini."
So, when it comes to Google, the word "Gemini can refer to the models, the subscription service, or the generative AI tools within Google Workspace and Cloud.
Oh, and if you’re wondering, Google’s Search Generative Experienced is still called SGE, but who knows, that could be rebranded to Gemini soon too.
There are two ways toaccess Gemini Ultra:
(BTW: Google’s currently offering Gemini Advanced for 2 months free.)
Just like to ChatGPT, Gemini will be available as a mobile app but only on Android devices for now. Mobile users will also have the option to set Gemini as their default assistant, replacing Google Assistant.
Though the company hasn’t announced any plans to phase out Google Assistant, it has reportedly been deprioritized, as AI becomes the company’s focus.
As for API access, developers and Cloud customers will have to wait a bit. Google is expected to share updates over the coming weeks.
The jury's still out on how Ultra compares to GPT-4, but it's safe to say OpenAI's got some solid competition for the first time in a while.