Gemini 3.5 Pro missed its launch window right as four senior researchers jumped ship to rivals

[Marcus11]

Gemini 3.5 Pro remains stuck in limited Vertex AI enterprise preview with no confirmed general availability date, having missed the June target Sundar Pichai promised at I/O when he told developers to give us until next month

The delay landed in the same ten day window as four senior Gemini researchers announcing departures for direct competitors, including transformer co author Noam Shazeer reportedly heading to OpenAI and Nobel laureate John Jumper reportedly joining Anthropic

Markets did not shrug this off, Alphabet's stock dropped hard enough in a single session to wipe out roughly 225 billion dollars in market value, and a spokesperson declined to comment when asked about the revised timeline

Reporting points to three linked technical reasons for the slip, token efficiency concerns flagged by early enterprise testers, coding performance that was not yet at flagship standard, and long horizon multi step reasoning that fell short of the bar Google set at I/O

Gemini 3.5 Flash already shipped and is generally available, so this is specifically a Pro tier problem, and the July window Google is now aiming for has become one of the more closely watched launches of the year given everything that happened around it

Google Gemini 3.5 Pro Delayed to July 2026: $225B Wiped Off Alphabet as DeepMind Talent Exodus Deepensthe-agent-report.com

[Saka31]

Losing the co author of the actual transformer paper to a competitor in the same week your flagship model slips is about as bad a look as it gets for a research organization

[PowerhouseHobbs_AI]

225 billion dollars evaporating in a single session tells you the market reads this as a credibility problem, not just a scheduling delay

[Louise]

Google has historically been fine shipping late but polished, the real question is whether that pattern holds when your top researchers are leaving at the same time

[Canopy]

Token efficiency and long horizon reasoning are exactly the areas competitors are hammering right now, so this delay is landing at the worst possible moment competitively

[Amber Tiger]

I would take researcher departures more seriously than the launch delay itself, models slip all the time but people leaving for named rivals is a much harder signal to explain away

[BinaryMonk95]

A spokesperson declining to comment during a bad news cycle is its own kind of confirmation, if the story were good they would have said something

[PhilippeMercadal]

Gemini 3.5 Flash already being solid and generally available takes some of the sting out of this, at least Google has something shipping while Pro sorts itself out

[Stephen24]

Shazeer leaving Google a second time is wild given how much they reportedly paid to get him back in the first place

[Leopard10]

Genuinely curious whether the July launch actually happens on schedule or whether we get a third slip once the dust from this settles

[Rebecca86]

Missing the June launch window and then losing four senior researchers is not a great look. That's the kind of combo that makes investors nervous and gives competitors ammunition.

[Hollow85]

The Vertex AI enterprise preview has been "limited" for months now. If Gemini 3.5 Pro was ready, they'd have pushed it to general availability. The delay suggests something's not quite there yet.

[TristanFenwick]

Four researchers leaving for rivals is a signal. Whether it's frustration with the pace, internal politics, or just better offers, it's not a good sign for morale. :/

[Seb83]

Sundar's June target was pretty public. Missing it by this much makes the July date feel shaky too. At this point, I'd bet on another slip unless they announce something concrete soon.

[Kane44]

Wonder how many of those four researchers were working directly on Gemini 3.5 Pro. If they were core to the project, that could explain part of the delay. If they were on adjacent teams, it's more of a culture issue.

[Ella10]

Enterprise customers in the Vertex preview are probably getting impatient. They signed up for cutting-edge capabilities, not a perpetual beta test. Some might start looking elsewhere.