 (3).png)
1 month ago
3
ARTICLE AD BOX
Salesforce CEO Marc Benioff has praised Google’s new Gemini 3, calling it a major leap in speed, reasoning and creative output. Early tests and benchmark wins suggest Google may have pulled ahead of rivals such as ChatGPT and Claude, as the AI race intensifies.
Google’s latest large language model, Gemini 3, has drawn striking praise from industry leaders after a high-profile early demo. Salesforce chief executive Marc Benioff said the model represented an extraordinary step forward in his post on Monday, adding that after only two hours with Gemini 3, he did not expect to return to ChatGPT. He highlighted faster performance, sharper reasoning and improvements across images and video.
Benioff also shared a Wall Street Journal report outlining how Google had finally surpassed its competitors with the new release.
Internal confidence backed by benchmarks
According to the report, the third version of Gemini has outperformed OpenAI’s ChatGPT and other leading models across standard industry benchmarks, giving Google a clear public win. Employees at the company had been testing the model privately for months and had already sensed a major breakthrough.
Tulsee Doshi, senior director of product management for Gemini, said one of her informal tests involved asking the system to write in Gujarati, a language widely used in India but less prominent online. The results, she said, showed a noticeable improvement compared with previous models.
Early access users report striking gains
Aaron Levie, chief executive of Box, said his company’s early evaluation of Gemini 3 delivered surprisingly strong results. After running tests on large and complex document sets, the cloud content management company found the model scoring significantly higher than expected. Levie said the performance jump was large enough that his team initially questioned whether they had run their assessments correctly.
Google claims long-sought advantage
The successful rollout marks a rare moment for Google, which has spent years trying to catch up in the AI race following ChatGPT’s explosive debut three years ago. Early reactions to Gemini 3 have highlighted its intelligence, accuracy and creative range.
The company has also said the model will power a new version of Nano Banana, an image-generation tool that has already helped drive a rise in Gemini use this year.
Michael Nathanson, an analyst at MoffettNathanson, said Google now appears to be firmly in the lead, calling the company clear winners in the latest round of AI development.
ChatGPT vs Claude vs Gemini
OpenAI retains the largest user base, with ChatGPT attracting around 800 million weekly users. Google, by contrast, reports 650 million monthly users for Gemini. Anthropic’s Claude also remains one of the preferred choices for coding-specific tasks. Even so, analysts say Gemini 3’s breadth of capability strengthens Google’s position across a wide range of use cases.
View full Image
Inside Google’s new AI strategy
The company has spent the past year restructuring its AI operations following concerns that it was losing ground. Chief executive Sundar Pichai and senior leaders have pushed for greater coordination across teams and consolidated model development. Co-founder Sergey Brin has also returned to a more active role, helping to guide research and product direction.
(With inputs from WSJ)
English (US) ·