alphabet is launching three new Gemini models on Tuesday, including the clearest answer yet Anthropoceneis a leader in cybersecurity as the company looks to show progress in a product pipeline that has faced delays and increasing competition.
Gemini 3.5 Flash Cyber is designed to detect and remediate software vulnerabilities and will initially only be available to governments and trusted partners as part of a limited-access pilot. Google said the special model runs at a lower price per token than larger models.
The new model could help Google close its cybersecurity gap over Anthropic, which gained an early lead in automated code defense with its Mythos model.
Google is also launching Gemini 3.6 Flash, which improves coding, multimodal and knowledge work performance while consuming up to 17% fewer tokens and costing less per token than the previous model – a significant reduction in the cost of running high-volume workloads.
Gemini 3.5 Flash-Lite is now Google’s fastest and most cost-effective model in the 3.5 family and is designed for high-volume workloads and smaller tasks within larger AI agent systems.
The broader offering reflects Google’s bet that price and efficiency can help offset slower timing in several key product categories.
Artificial analysis data shows that Gemini Flash is already outperforming comparable models from Anthropic. OpenAI and Chinese competitors on costs. According to the company, Gemini 3.6 Flash – the most powerful of Alphabet’s new models – is cheaper per task than GPT-5.6 Terra Max, Kimi K3 and Qwen 3.7 Max.
The launch comes on the eve of Alphabet’s results and as its Chinese rivals gain momentum. Moonshot AI’s Kimi K3 attracted so much demand that the company limited new subscriptions and API access due to capacity constraints, while Alibaba is testing Qwen 3.8 Max, which it says trails only Anthropic’s Fable 5 in overall performance.
This demand illustrates the other side of the AI race: building a competitive model is only part of the challenge. Companies also need enough computing capacity to deploy on a large scale.
Google has a potential advantage through its custom chips, cloud infrastructure and ability to co-develop models and hardware, although the company itself has faced capacity constraints.
Tuesday’s model launches come as Google is reportedly developing a special chip that will make Gemini up to 10 times more efficient. This is part of a broader initiative to reduce the cost of deploying AI.
A Google Cloud spokesperson told CNBC in a statement that its teams “continuously explore and experiment with new innovations to deliver maximum performance and efficiency to our users and customers” and that “while not every project moves into production, this rigorous exploration is central to our full-stack approach.”
“By co-developing our hardware and software from the ground up, we ensure our systems are integrated and highly optimized for real-world workloads,” the statement continued.
Google is also offering more insight into its roadmap following questions about delays. Gemini 3.5 Pro is now being tested with partners before it becomes more widely available, while the company has begun the largest pre-training run for Gemini 4 to date.
REGARD: Dow achieves record close with Alphabet as new driver
Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.
https://www.cnbc.com/2026/07/21/google-gemini-flash-ai-mythos-rival.html
