Google’s Gemini 3 is winning over tech CEOs, researchers: Here’s what they are saying

Gemini 3 has also been designed to focus on the context and intent behind user requests in order to reduce the need for extensive prompting. In a first for the company, the AI model has also been integrated into Google’s search products from the start, a move that likely signals the company’s confidence in Gemini 3’s multimodal reasoning capabilities.
However, the Gemini 3 roll-out has not been without its gaffes. Let’s take a look at some of the notable reactions that have followed Google’s release of its most advanced AI model yet.

‘Not going back to ChatGPT’

Salesforce founder Marc Benioff said that while he has used OpenAI’s ChatGPT every day for the past three years, he is making the switch to Gemini 3. “Just spent 2 hours on Gemini 3. I’m not going back. The leap is insane — reasoning, speed, images, video… everything is sharper and faster. It feels like the world just changed, again,” Benioff wrote in a post on X.

In October this year, Benioff and Google CEO Sundar Pichai shared the stage at Dreamforce 2025. “You’re from my favourite part of India, which is the south,” Benioff told Pichai. The two tech giants also announced an expansion of their strategic partnership which includes integrations of Google’s cutting-edge Gemini models in Salesforce’s new Agentforce 360 Platform.

‘A major leap in reasoning’

Aaron Levie, the CEO of cloud content management platform Box, said that his company was given early access to Gemini 3 Pro which is said to have better long-horizon planning to generate significantly higher returns compared to other frontier models.

Gemini Pro is a massive 22 percentage point improvement over Gemini 2.5 Pro, according to Levie. “Gemini 3 Pro represents a major leap in reasoning, math, logic, and analytical capabilities. These gains show up as delivering highly useful improvements across financial services, law, healthcare, public sector, manufacturing, and more,” he wrote in a post on X.

‘Clearly a tier 1 LLM’

Famed AI researcher Andrej Karpathy said that he was given early access to Gemini 3, and his impression of the model was positive in terms of its personality, humour, writing skills, and vibe-coding abilities.

“Very solid daily driver potential, clearly a tier 1 LLM, congrats to the team!” Karpathy wrote in a post on X. However, he also pointed out that Gemini 3 refused to believe that it was 2025 since its pre-training data only included information up till 2024.

He also warned that public benchmarks are unreliable indicators of real-world AI performance because they can be easy to game. “Over the next few days/weeks, I am most curious and on a lookout for an ensemble over private evals, which a lot of people/orgs now seem to build for themselves and occasionally report on here,” Karpathy said.

‘Moves the frontier forward’

Gemini 3 has been termed as a “genuinely strong model” and “the current state-of-the-art” by investors and analysts. “We’d go as far to say that this latest model from [Google DeepMind] meaningfully moves the frontier forward, with capabilities that in certain areas far exceed what we’ve typically come to expect from this generation of frontier models,” analysts at DA Davidson were quoted as saying by CNBC.

“Healthy adoption metrics for AI Overviews and Gemini indicate Google is successfully funneling users into its AI surfaces, despite growing competition, and should help ease concerns around potential search disruption,” Bank of America Securities analysts reportedly said, adding that Gemini 3 represents “another positive step” for Google to close any “perceived LLM performance gap” to its AI competitors.

Meanwhile, Google parent Alphabet’s stock jumped three per cent the day after Gemini 3 was announced. Its shares are up more than 55 per cent so far this year, as per the report. Additionally, Warren Buffett’s Berkshire Hathaway recently disclosed a new stake in the Google parent, marking one of the conglomerate’s most significant technology investments in years.