Google is expanding the Gemini lineup with cheaper models and new Mythos competitors
Alphabets releases three versions of Gemini on Tuesday, including its clearest response to date Anthropic‘is at the forefront of cybersecurity, as the company looks to demonstrate progress across its product line in the face of delays and increased competition.
Gemini 3.5 Flash Cyber is designed to identify and patch software vulnerabilities and will only be available to governments and trusted partners through a limited access driver. Google said that the special model works at a lower price per token than the main models.
That could help Google narrow its cybersecurity gap with Anthropic, which has built itself at the forefront of automated code protection.
Google also introduced Gemini 3.6 Flash, which improves coding, multimodal performance and information work while using up to 17% fewer tokens and costs less per token than the previous model – a logical decrease in the cost of using large volumes.
Gemini 3.5 Flash-Lite, on the other hand, is Google’s fastest and most cost-effective model in the 3.5 family, designed for large workloads and small tasks within large AI agent systems.
The broad list reflects Google’s bet that price and efficiency can help offset its slow pace in several key product categories.
Artificial intelligence data shows that Gemini Flash already undercuts comparable models from Anthropic, OpenAI and Chinese competitors on costs. According to the company, the Gemini 3.6 Flash – powered by these two new models – is cheaper per operation than the GPT-5.6 Terra Max, Kimi K3 and Qwen 3.7 Max, while the 3.5 Flash-Lite costs a fraction of that.
The release comes on the eve of Alphabet’s acquisition and as Chinese rivals gain momentum. Moonshot AI’s Kimi K3 drew enough demand that the company limited new subscriptions and API access due to power limitations, while Alibaba teased the Qwen 3.8 Max, saying it trails only Anthropic’s Fable 5 in overall performance.
That demand highlights another side of the AI race: Building a competitive model is only part of the challenge. Companies also need sufficient computing capacity to operate at scale.
Google has a potential advantage with its custom chips, cloud infrastructure and ability to design models and hardware together, although the company has faced its own problems.
The launch of Tuesday’s model comes as Google is reportedly developing a special chip designed to run Gemini 10 times more efficiently, part of a broader plan to lower AI costs.
A Google Cloud spokesperson told CNBC in a statement that its teams are “constantly researching and testing new things to bring efficiency and effectiveness to our users and customers” and that “while not all projects make it to production, this rigorous testing is the foundation of our holistic approach.”
“By integrating our hardware and software from the ground up, we ensure that our systems are integrated and highly optimized to handle real-world workloads,” the statement continued.
Google also provides more visibility into the roadmap behind queries about delays. The Gemini 3.5 Pro is now undergoing testing with partners ahead of wider availability, while the company has started its biggest-ever training run for the Gemini 4.
WATCH: Dow closes records with Alphabet as new driver
