It was only a few days ago we heard rumors around OpenAI launching a next-gen ChatGPT-5 model – rumors dismissed as “fake news” by OpenAI chief Sam Altman – and now we’re hearing Google Gemini 2.0 could be made public in the next couple of months.
Like the ChatGPT rumor, this comes from The Verge (via 9to5Google). The last major Gemini upgrade came back in February, when Gemini 1.5 was pushed out. Google uses the Gemini name for both its AI bots, and the underlying models.
What we don’t get in this report are any indications of what might be new and improved in Gemini 2.0. If previous releases are anything to go by, expect smarter answers, faster processing, support for longer inputs, and more reliable reasoning and coding.
As per the report, Gemini 2.0 doesn’t offer the sort of performance increases its developers were originally hoping for – but the article also notes that this is a trend affecting all large language models (LLM) in general, not just those at Google.
Keeping up the pace
If Google (and perhaps OpenAI) did indeed launch AI model upgrades before the end of the year, it would be further evidence of the time and resources that companies are investing in artificial intelligence in order to stay ahead.
In recent weeks we’ve seen new video tools, improved image generation, and bespoke search features launched by AI companies. When next week rolls around, meanwhile, millions of iPhone owners will be able to try out Apple Intelligence for the first time..
That said, we’re still waiting for Google to launch the next-gen Project Astra AI assistant we saw demoed at Google I/O 2024 back in May. Project Astra combines multiple inputs and outputs in a more advanced and natural way than ever.
While we have previously spent some hands-on time with Project Astra, it doesn’t seem as though it’s directly connected to Gemini 2.0 – though it may be powered by it. Of course as soon as anything is made official, we’ll bring you the news here.