[DigitalToday reporter Chi-gyu Hwang (황치규)] Google has unveiled 3 new Gemini models that highlight price and efficiency. It also released a security-focused model. CNBC reported on July 21 that Alphabet launched Gemini 3.5 Flash Cyber, Gemini 3.6 Flash and Gemini 3.5 Flash-Lite.
Gemini 3.5 Flash Cyber is a model that detects and patches software vulnerabilities. At first, only governments and trusted partners will be able to use it in a restricted-access pilot. Google explained that the model has a lower per-token price than larger models.
Gemini 3.6 Flash improved performance for coding, multimodal and knowledge work. It uses up to 17 percent fewer tokens than the previous model and also reduced cost per token. Gemini 3.5 Flash-Lite is the fastest and cheapest model in the 3.5 line, and is suitable for batch work and smaller tasks within large AI agent systems.
Artificial Analysis said Gemini Flash already costs less than similar models offered by Anthropic, OpenAI and Chinese competitors. Google said Gemini 3.6 Flash has a lower cost per task than GPT-5.6 Tera Max, Kimi K3 and Qwen 3.7 Max, and that 3.5 Flash-Lite is much cheaper than that.