Google is expanding its Gemini portfolio with three new AI models designed to improve performance, reduce costs, and support production AI deployments.
The latest releases, Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber, target enterprise use cases ranging from coding and AI agents to document processing and cybersecurity. Each model is designed for a different workload, giving organizations more flexibility when deploying AI.
Gemini 3.5 Pro did not debut alongside the new Flash models despite many people expecting Google’s next flagship model to arrive with this release.
Why It Matters: Organizations are no longer looking for one AI model that does everything. They want models that fit specific workloads while controlling costs and supporting production deployments. Google’s latest additions expand the Gemini portfolio with that goal in mind, giving organizations more flexibility as AI adoption grows across business operations.
- Gemini 3.6 Flash Takes Center Stage: Google says Gemini 3.6 Flash delivers stronger coding performance while using up to 17% fewer output tokens than Gemini 3.5 Flash. The lower token usage helps reduce inference costs, making the model more economical to run in production. Google also improved latency and reliability, with a focus on organizations building AI agents for day-to-day business use.
- Two Models Target Specialized Workloads: Gemini 3.5 Flash-Lite is Google’s most cost-effective model in its class. It is designed for high-volume tasks such as AI agents and document processing, where keeping costs low is a priority. Gemini 3.5 Flash Cyber focuses on identifying and remediating software vulnerabilities. Google will initially make the cybersecurity model available only to governments and trusted partners through a limited-access CodeMender pilot.
- Gemini Expands Its Portfolio: Instead of asking one model to handle every workload, Google is giving customers more choice. Some organizations may prioritize affordability for high-volume tasks, while others may need stronger coding capabilities or a dedicated cybersecurity model. Organizations can select models based on the needs of each application, whether the priority is speed, security, or lower operating costs.
- Gemini 3.5 Pro Remains on Hold: Google previously indicated the next-generation Pro model would arrive shortly after the Flash releases, yet it remains in partner testing following reported internal delays tied to performance goals. Google says it will release Gemini 3.5 Pro once it is ready. The company also revealed that Gemini 4 has entered its most ambitious pre-training run to date.
- Competition Continues to Intensify: Since Google’s last Pro model update in February, OpenAI and Anthropic have continued introducing new frontier models. Competition now extends beyond model capability to include pricing, deployment efficiency, and specialized offerings. Google is also working to retain top DeepMind researchers during an aggressive recruiting battle across the AI industry. For enterprise customers, the result is a wider selection of AI models designed for different business needs.
Go Deeper -> Google releases three new Gemini models — but no 3.5 Pro – Tech Crunch
Google releases series of new cheaper Gemini models – AXIOS
Trusted insights for technology leaders
Our readers are CIOs, CTOs, and senior IT executives who rely on The National CIO Review for smart, curated takes on the trends shaping the enterprise, from GenAI to cybersecurity and beyond.
Subscribe to our 4x a week newsletter to keep up with the insights that matter.
