Google Releases Three New Gemini Models, but Flagship Model Still Delayed
Miles Bennett
Google released three new Gemini models on Tuesday, all built around speed and lower cost, but the long-awaited flagship Gemini 3.5 Pro has been delayed again — Google is playing the efficiency card to buy time, at the price of dropping out of the top performance tier.
What role does each new model play?
Gemini 3.6 Flash is positioned as Google's workhorse — it outperforms the previous 3.5 Flash on coding tasks, and token consumption (the billing unit for model calls) is down up to 17%.
Gemini 3.5 Flash-Lite is the fastest and cheapest in the 3.5 family — built for high-volume, cost-sensitive workloads.
Gemini 3.5 Flash Cyber targets cybersecurity vulnerability detection and patching. Google says performance matches rivals, but per-token pricing undercuts larger models.
Why do all three emphasize "cheaper"?
The pricing strategy takes direct aim at Anthropic and OpenAI — both recently launched cybersecurity models. Google chose to compete on cost, not raw performance.
Google CEO Sundar Pichai spelled out the logic: "Companies are already burning through their annual token budgets, and it's only May."
This means → Google is betting that enterprise customers' core pain point is not "who's strongest" but "can the budget last through December." Mixing Flash models with frontier models saves money.
Why does the flagship Gemini 3.5 Pro keep slipping?
3.5 Pro was originally set for June. Business Insider reported the target moved to July; Bloomberg recently reported it may slip further.
On Tuesday Google said only that 3.5 Pro is being tested with partners and will ship "as soon as it's ready" — no new date.
This reflects a development timeline that has drifted from Google's original roadmap. The company has not explained the delays.
What does the flagship gap mean?
As of Tuesday, Google holds no position in the top ten on Artificial Analysis's composite benchmark — covering math and reasoning.
This means → until 3.5 Pro or the next-generation Gemini 4 lands, Google is on the back foot in the frontier-performance race — relying on efficiency and price alone to hold enterprise clients.
Google announced the same day that it has begun "the most ambitious Gemini 4 pre-training to date," but that model is still months away. In plain terms = Google has cards to play, but the biggest card isn't on the table yet — the gap in between is the real risk.
Content is for reference only, not financial advice.