In the AI race, the throne is never safe. Just ask Google.
After the success of Gemini 3, Google found itself in a strong position at the end of 2025. As of this week, the situation is a little shakier.
While the company just rolled out three faster, more cost-effective models, it continues to delay its next frontier model, Gemini 3.5 Pro, and it’s unclear whether this week’s launches will be enough to keep users and investors happy in the interim.
Some of Google’s competitors are using the opportunity to land a few jabs.
Alexandr Wang, Meta’s chief AI officer, wrote on X “gemini who?” in response to a leaderboard that ranked Meta’s Spark model above one launched by Google this week.
Thibault Sottiaux, a member of technical staff at OpenAI, also took an apparent jab at the search giant. In a post on X, Google’s Logan Kilpatrick announced that pre-training on Gemini 4 — the next big milestone model — had begun. “Hope it finishes one day too!” Sottiaux replied.
Google declined to comment.
‘Too early to count anyone out’
Google’s delay is particularly glaring because OpenAI and Anthropic have rolled out new top-tier models in recent weeks. The pushback of Gemini 3.5 Pro has “shifted perception from leading edge to trailing edge,” said Josh Beck, an analyst at Raymond James, in a note this week. He said he saw this as a byproduct of the fast pace of change among the labs right now.
At the same time, Google’s business has been humming along nicely in recent quarters, with strong momentum across Search, YouTube, Cloud, and other areas benefiting from Google’s AI advancements. Google is also betting that faster, more cost-effective models may be a winning strategy at a time when token costs are racking up.
Google’s focus on more efficient models has received praise from some users.
“Google gets a lot of criticism on here for falling behind on agentic coding, but Gemini 3.5 Flash has long been my daily driver for agentic document extraction, which is one of the highest-value use-cases for LLMs IMO,” Kyle Walker, founder of Clearfork Intelligence, wrote on X.
Still, Google may need to address this trade-off between efficiency and power when it announces Q2 earnings on Wednesday evening. Analysts are likely to raise the topic of 3.5 Pro and its release timeline.
“I love Gemini, probably more than I should but them hyping 4 before even delivering 3.5 Pro is a lil weird,” Anshel Sag, analyst at Moor Insights & Strategy, wrote on X.
Sag told Business Insider he felt that Google hyping up Gemini 4 was an “admission they already have something better.” However, he said the “feverish pace” of AI right now doesn’t necessarily yield meaningful improvements.
“I just feel like Google is a much bigger company and moves a bit differently from its competitors,” said Sag.
He added: “It’s just way too early to count anyone out.”
Have something to share? Contact this reporter via email at hlangley@businessinsider.com or Signal at 628-228-1836. Use a personal email address and a non-work device; here’s our guide to sharing information securely.
