Google Revolutionizes AI with Gemini 3.6 Flash and Lower Prices

Google Unveils Its New Gemini Flash Models

Google continues to expand its range of artificial intelligence models under the Gemini banner. In an announcement made on July 21, 2026, the company revealed three new models designed to optimize the execution of large-scale AI agents. The Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber models all aim to provide greater efficiency, reduced latency, and lower costs.

A Series Focused on Efficiency and Cost Reduction

The Flash series of Gemini is designed to balance quality and efficiency in processing agent workflows. The three new models fit into this strategy:

• Gemini 3.6 Flash is considered the flagship model of the series, offering enhanced performance in development, knowledge management, and multimodal applications. According to the Artificial Analysis Index, it reduces output token consumption by 17% compared to 3.5 Flash, with improvements of up to 65% on certain benchmarks such as DeepSWE.

• Gemini 3.5 Flash-Lite is the fastest and most cost-effective model in the 3.5 series. It can generate 350 tokens per second, making it ideal for high-volume tasks such as agent-based research and document processing.

• Gemini 3.5 Flash Cyber is a specialized model designed to identify and correct security vulnerabilities. It is integrated into CodeMender, Google's security agent.

The first two models, Gemini 3.6 Flash and 3.5 Flash-Lite, are now available for developers via the Gemini API, for businesses through Gemini Enterprise, and for the general public in the Gemini app. The Cyber model, due to its dual-use potential, is reserved for governments and trusted partners as part of a restricted access pilot program.

Pricing for the New Flash Models

For one million tokens, the pricing is as follows:

• Gemini 3.6 Flash: $1.50 for input and $7.50 for output.

• Gemini 3.5 Flash-Lite: $0.30 for input and $2.50 for output.

Google emphasizes that the 3.6 Flash model is offered at a lower price than its predecessor, the 3.5 Flash.

The Notable Absence of Gemini 3.5 Pro

While these new models are diverse, the absence of a performance-focused version is notable. Google has yet to update its premium model, Gemini Pro, which last saw an update in February 2026. During the launch of Gemini 3.5 Flash last May, Google had announced the imminent arrival of a 3.5 Pro.

In its statement, Google clarified that 3.5 Pro is currently in testing with partners and will be deployed "as soon as it is ready." According to Bloomberg, the company is facing internal delays, with the model struggling to meet its performance targets, particularly in coding. Meanwhile, teams have launched the most ambitious pre-training cycle to date for Gemini 4.

This timeline comes amid intense competition. Since February, OpenAI has launched GPT-5.5 and then GPT-5.6, while Anthropic has introduced Claude Opus 4.8, Sonnet 5, and expanded access to its Fable 5 model.