New Delhi: Google has recently announced and unveiled its latest three AI models aimed at speed, lower costs, and enterprise use cases instead of releasing its long-awaited flagship models: Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber.
- Gemini 3.6 Flash: It is now faster and more capable than previous Flash models, with improvements in coding, reasoning, multimodal understanding, and agent workflows.
- Gemini 3.5 Flash-Lite: Google’s cheapest and fastest model in the Gemini 3.5 family, which is specially designed for high-volume apps like chatbots, customer support, document processing and summarization.
- Gemini 3.5 Flash Cyber: Specialized cybersecurity model which is built to detect software vulnerabilities, analyze source code, and recommend security fixes. It is available only to government and trusted partners.
Google has planned to unveil its Gemini 3.5 Pro in June 2026, but it has now slipped again. According to reports, the model hasn’t met Google’s internal quality aims. Coding performance is reported to be the biggest challenge for the company.
Enterprise customers heavily judge AI models on programming and software engineering tasks. Google stated that the model remains in experimentation and will debut soon.
Google appears to be shifting towards lower inference costs, faster response time, specialized AI models, and better enterprise ROI. Sundar Pichai emphasized that reducing AI deployment costs is becoming just as important as improving model intelligence.
Even as Gemini 3.5 Pro remains unreleased, Google has also confirmed it has begun the largest pre-training run yet for Gemini 4, suggesting the next-generation flagship is already under active development.
