Google Releases Gemini 3.7 Flash, Flagship Pro Version Still Has No Timeline
Nashnova编辑部
Google released Gemini 3.7 Flash on August 13, a low-cost coding model priced at half its predecessor, but the flagship Gemini 3.5 Pro that investors watch most closely still has no release date — leaving Google's timeline for closing the gap with Anthropic and OpenAI undefined.
What does Gemini 3.7 Flash actually bring?
Arriving just three weeks after 3.6 Flash, the new model targets coding tasks: debugging, bug fixes, and production-grade code generation.
Pricing is cut sharply: $0.75 per million input tokens, $3.75 per million output tokens — roughly half the 3.6 Flash price. This means → Google is waging a price war for developer adoption, prioritizing volume over margin.
The model is live on Gemini Spark across 160+ countries for AI Pro and Ultra subscribers.
Why is the flagship Pro version still missing?
Investors treat Gemini 3.5 Pro as the key benchmark for whether Google DeepMind can catch its rivals. Google said in July it would come "soon" — no date has followed.
Reuters reported that Google has cancelled the 3.5 Pro launch; DeepMind product lead Logan said Gemini 4 pre-training has begun, and a direct jump to Gemini 4 is possible. In plain terms = 3.5 Pro may never ship — Google may skip an entire generation and bet on the next one.
Morningstar analyst Malik Ahmed Khan said Anthropic and OpenAI lead Google in coding capability by "more than one full step," while Google directed more resources toward search monetization and multimodal work, missing the coding use case — the first to produce a breakout application.
How is leadership responding to the "falling behind" narrative?
CEO Sundar Pichai denied Google is behind on the July earnings call, defending the company's AI strategy.
But Reuters reported exclusively that co-founder Sergey Brin has recently urged core AI staff to go all-in on Gemini to help parent company Alphabet close the gap. This reflects a level of internal urgency that may exceed the public messaging.
What do the leadership changes and restructuring signal?
One week before the launch, Google carried out a major leadership reshuffle at DeepMind: co-founder Demis Hassabis stepped down, replaced as senior VP by his deputy Koray Kavukcuoglu, who reports directly to Pichai.
The two original technical co-leads of the Gemini project also departed and co-founded a startup. This means → the founding team behind Google's flagship AI project has largely turned over; whether the new leadership can deliver is an open question.
Constellation Research chief analyst Ray Wang said Kavukcuoglu's promotion signals Google is putting execution ahead of research prestige, potentially accelerating product release cycles.
Can Flash's rapid iteration solve the real problem?
Organizationally, several non-technical teams have been moved out of DeepMind and now report through Google's corporate structure; the research team stays within DeepMind. In plain terms = Google is splitting "the people who ship products" from "the people who do research," aiming to push Gemini into Search, Google Cloud, and other product lines faster.
Malik Ahmed Khan said that under the new leadership, "improving LLM capability will become a clear path forward."
But the core question remains: Flash's three-week iteration cycle proves Google's engineering speed; whether the flagship model can deliver a genuine breakthrough is the real test of whether Google's AI competitiveness can recover.
Content is for reference only, not financial advice.