Google DeepMind launches new AI models emphasizing efficiency

Google DeepMind launches new AI models emphasizing efficiency

Google DeepMind has introduced three new proprietary AI models: Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber, which the company states are among its most token-efficient to date.

The models are designed to make AI agents faster, smarter, and more cost-effective at scale. Gemini 3.6 Flash is priced at $1.50 per million input tokens and $7.50 per million output tokens via its API. Gemini 3.5 Flash-Lite is offered at a significantly lower rate of $0.30 and $2.50 per million input and output tokens, respectively.

These new models offer considerable cost savings compared to previous iterations and some competitors, particularly for tasks requiring extended processing. While the older Gemini 3.1 Flash-Lite remains the most cost-efficient, the new Gemini 3.5 Flash-Lite offers twice the speed, providing businesses that prioritize performance a more economical option.

Gemini 3.6 Flash and Gemini 3.5 Flash-Lite are immediately available through the Gemini API. Gemini 3.5 Flash Cyber is tailored for cybersecurity research and will be accessible to select governments and partners via the CodeMender platform. All models are proprietary and closed-source, accessible only through Google's official APIs.