In the accelerating contest for artificial intelligence supremacy, Google has released Gemini 3.7 Flash — a leaner, cheaper model aimed at developers and automated workflows — while conspicuously withholding its more powerful Pro counterpart. The launch, arriving just three weeks after its predecessor, speaks less to confident momentum than to the particular urgency of a company reorganizing itself mid-race. What Google offers today is speed and affordability; what the market is still waiting for is proof that it can lead.
Google launches Gemini 3.7 Flash for coding, delays premium Pro model
Google needs to prove it can match them at the frontier
Mark
Why does the timing of Gemini 3.5 Pro matter so much to investors?
Mimi
Because it's the test case. If Google can't deliver a competitive premium model, it suggests the company's AI research engine isn't moving fast enough. Anthropic and OpenAI have already shown what's possible at the top end. Google needs to prove it can match them.
Mark
But Google just released a new Flash model with better performance. Isn't that a win?
Mimi
It is, but it's a different kind of win. Flash is the workhorse—cheaper, faster, good enough for most coding tasks. But it's not the frontier. The Pro model is where you prove you're still innovating at the highest level.
Mark
The leadership changes at DeepMind seem significant. What's really happening there?
Mimi
Restructuring under pressure. Hassabis stepping back, two technical leads leaving to start a company—these are signals that something wasn't working the way it needed to. Google is trying to reset and move faster.
Mark
Is Pichai's defense of the strategy credible?
Mimi
He's defending a strategy that hasn't yet proven itself. The Flash releases are real, the pricing is aggressive, the reorganization is real. But until Pro ships and performs, the defense is incomplete.
Mark
What does Google need to happen in the next few months?
Mimi
Gemini 3.5 Pro needs to land and it needs to be genuinely competitive. Not just good—genuinely competitive. And it needs to happen soon enough that people believe Google's AI research is still moving forward, not playing catch-up.
El Pulso
- Google's silence on Gemini 3.5 Pro — promised in July and still absent in August — is being read by investors and rivals as a sign of genuine strain in the premium AI race.
- Internal tremors are visible: DeepMind's founding leader Demis Hassabis has stepped aside, two of Gemini's original technical co-leads have left to start their own venture, and co-founder Sergey Brin is reportedly pressing staff to commit fully to the effort.
- To hold ground while the flagship model is delayed, Google is pricing Gemini 3.7 Flash aggressively — half the cost of its predecessor — and pushing it through its subscription service to customers across more than 160 countries.
- CEO Sundar Pichai has publicly defended Google's AI standing, but the company's own actions — rapid release cycles, steep price cuts, leadership reshuffles — tell the story of an organisation in pursuit, not in command.
In the accelerating contest for artificial intelligence supremacy, Google has released Gemini 3.7 Flash — a leaner, cheaper model aimed at developers and automated workflows — while conspicuously withholding its more powerful Pro counterpart. The launch, arriving just three weeks after its predecessor, speaks less to confident momentum than to the particular urgency of a company reorganizing itself mid-race. What Google offers today is speed and affordability; what the market is still waiting for is proof that it can lead.
Google released Gemini 3.7 Flash on Thursday, marketing it as a practical tool for software developers and businesses building autonomous systems capable of handling complex, multi-step tasks with little human intervention. The model arrives just three weeks after its predecessor and shows real improvements in coding performance — writing production-ready code, debugging, and real-time problem-solving. Yet the announcement was defined as much by what was missing: no timeline for Gemini 3.5 Pro, the flagship premium model that investors and competitors have been waiting on since Google described it as coming "soon" back in July.
To keep customers engaged in the interim, Google is pricing Flash aggressively — 75 cents per million input tokens through year-end, half what the previous version cost — and distributing it through its Gemini Spark subscription service in more than 160 countries. The strategy is clear: make Flash indispensable while the more consequential model is still being prepared.
Behind the product announcements, Google's AI organisation is visibly in flux. Demis Hassabis has stepped back from leading DeepMind in favour of his deputy, Koray Kavukcuoglu. Two of Gemini's original technical co-leads have departed to found their own company. Co-founder Sergey Brin has reportedly been pressing key staff to commit more fully to the Gemini effort. CEO Sundar Pichai defended the company's position during its July earnings call, but the pattern of rapid releases, price reductions, and leadership changes points to an organisation working hard to close a gap rather than extend a lead.
The question that will define Google's near-term standing in AI is simple: when does Gemini 3.5 Pro ship, and what can it actually do against what Anthropic and OpenAI have already built? Until that answer arrives, Flash will carry the weight — capable, affordable, and available now. But availability is not the same as dominance.
Google rolled out Gemini 3.7 Flash on Thursday, positioning it as a streamlined tool for software developers and companies building autonomous systems that can handle multi-step workflows with minimal human oversight. The model arrives three weeks after its predecessor, Gemini 3.6 Flash, and shows measurable gains in the kinds of tasks that matter most to engineers: writing production-ready code, debugging, and fixing problems in real time. But the announcement carried an absence as loud as any presence—Google offered no timeline for releasing Gemini 3.5 Pro, its flagship premium model, leaving investors and competitors watching closely for signs of whether the company can actually keep pace with Anthropic and OpenAI in the race for advanced AI capabilities.
The timing reveals something about Google's current position. Back in July, the company had promised that Gemini 3.5 Pro was in testing with partners and would arrive "soon." That word has now stretched into August with no update. Meanwhile, the company is betting on Flash as a workhorse for the near term, pricing it aggressively to win adoption. Through the end of the year, businesses can access Gemini 3.7 Flash at 75 cents per million input tokens and $3.75 per million output tokens—half what Google charged for the previous version. The model is already available through Gemini Spark, Google's subscription AI agent service, which reaches customers in more than 160 countries who pay for Google AI Pro or Ultra tiers.
The delay in the Pro model matters because it signals something about the pressure Google faces. Sergey Brin, the company's co-founder, has been pushing key AI staff internally to commit fully to the Gemini effort, according to reporting from Reuters. That urgency reflects a real competitive gap. Last week, Google announced a significant restructuring of its DeepMind division: Demis Hassabis, who led the unit, stepped back from the chief role in favor of his deputy, Koray Kavukcuoglu. At the same moment, two of Gemini's original technical co-leads departed to start their own company. These are not the moves of an organization coasting.
CEO Sundar Pichai defended Google's AI strategy during the company's July earnings call, pushing back against concerns that Google had fallen behind and lost ground in AI-powered coding. His defense was vigorous, but the facts on the ground tell a more complicated story. The company is accelerating its release cycle—three weeks between Flash versions is a sprint—and cutting prices to drive adoption. It is also reorganizing leadership and watching talented people leave. These are the actions of a company trying to catch up, not one that has already arrived.
What comes next will likely depend on when Gemini 3.5 Pro actually ships and what it can do. The market is waiting to see whether Google's DeepMind unit, even with new leadership and renewed focus, can deliver a premium model that genuinely matches what Anthropic and OpenAI have built. Until then, Flash will carry the load—fast, affordable, and available now. That may be enough to keep customers engaged. But it is not the same as leading.
Citas Notables
Google co-founder Sergey Brin urged key AI staff to commit fully to the Gemini model as Alphabet seeks to close the gap with rivals— Reuters reporting
CEO Sundar Pichai defended Google's AI strategy during earnings, pushing back on concerns the company has fallen behind rivals— Pichai, during July earnings call
Quieres la nota completa?
Lee el original en The Hindu ↗
Technology