In its ongoing effort to maintain its competitiveness in the advanced technology market, Google is adopting a strategy of rolling out specialized, high-speed models that efficiently meet developers’ needs at a low cost.
- Launch of three new models focusing on speed and efficiency
- Gemini 3.6 Flash as an ideal choice for work tasks
- Exceptional speed of the new Gemini 3.5 Flash Lite model
- Adapting artificial intelligence in the field of cybersecurity
- Delay of the flagship Gemini 3.5 Pro release and its technical reasons
- Targeted efficiency strategy and meeting the demands of AI agents
- Frequently asked questions
Launch of three new models focusing on speed and efficiency
In a strategic move aimed at enhancing its portfolio of technology solutions and providing diverse options for developers, global company Google launched three new smart models belonging to the Gemini family on Tuesday. These models include Gemini 3.6 Flash, Gemini 3.5 Flash Lite, and Gemini 3.5 Flash Cyber. In its announcement, the company strongly emphasized the factors of ultra-high speed and cost-efficiency, targeting these releases at developers working on building autonomous software agents that require fast response times. However, this launch fills a noticeable gap in the technical landscape amid the ongoing delay of the release of its flagship and most anticipated model, Gemini 3.5 Pro, which was scheduled to be released to the public several months ago but remains unavailable to this moment.
Gemini 3.6 Flash as an ideal choice for work tasks
According to the detailed post shared by the company on its official blog, the new models were introduced as specialized tools designed for specific purposes rather than comprehensive general models. The company described the Gemini 3.6 Flash model as the “backbone” or the ideal and most powerful choice for demanding daily work, delivering significant and substantial improvements compared to its direct predecessor in terms of coding, cognitive work tasks, and multimodal data processing. According to Artificial Analysis, a specialized index evaluating algorithmic performance, this new model succeeded in reducing output token consumption rates by up to 17 percent compared to the previous 3.5 Flash model, saving massive operational costs for companies and developers using it intensively.
Exceptional speed of the new Gemini 3.5 Flash Lite model
At the same time, the Gemini 3.5 Flash Lite model features unmatched, exceptional speed, possessing the superior ability to process and generate up to 350 tokens per second. This tremendous speed makes it the fastest 3.5-class model Google has ever released to date, and it is specifically designed to effectively handle massive and large-volume workloads where low cost and speed are the decisive factors for project success. Both the 3.6 Flash and Flash Lite models are available and open for use starting Tuesday through APIs in the company’s AI Studio, the enterprise-focused agent platform, and the consumer-facing app, to meet all needs at unprecedented speeds.
Adapting artificial intelligence in the field of cybersecurity
As for the third model, Gemini 3.5 Flash Cyber, it represents a new and distinct qualitative direction; it has been customized and trained exclusively to serve the sensitive cybersecurity sector. This model works side by side and in full integration with the company’s software security agent called Code Minder, and its primary mission is to search and hunt for security vulnerabilities in code and work to automatically and immediately patch and fix them to prevent breaches. Unlike the other two models available to the general public, journalistic sources, such as a report by Business Insider, indicated that this precise security model will initially be available exclusively to government agencies and highly trusted partners to ensure the safe and responsible use of this advanced technology in protecting critical infrastructure from cyberattacks.
Delay of the flagship Gemini 3.5 Pro release and its technical reasons
These new and fast releases come against a murky and disappointing backdrop due to the long and ongoing delay of the flagship Gemini 3.5 Pro model. Company CEO Sundar Pichai told journalists and media outlets ahead of the annual developer conference last May that the most advanced model would be “rolled out to everyone next month,” pointing clearly and directly to June, yet that deadline has already passed without delivering the promised product. Bloomberg News reported last week in an analytical report that the reason for this fundamental delay is the model’s failure to reach the expected levels and standards of the company’s internal goals, specifically regarding the performance and efficiency of writing and generating complex code.
Targeted efficiency strategy and meeting the demands of AI agents
The company acknowledged these technical challenges and delays in a brief official statement to 9to5Google, confirming that it is working hard and currently testing version 3.5 Pro and other advanced models in cooperation with its commercial partners and the United States government to ensure the highest standards of quality and safety. Tuesday’s official post did not provide any new timeline or specific, reliable date for the launch of the awaited professional version. This strategy reflects a clear direction by Google focusing on operational efficiency rather than sheer power, as Tulsi Doshi, an executive in the product management department, explained that the new releases represent the ideal and optimal blend of efficiency and power to run autonomous agents effectively, while the question remains open as to developer satisfaction with this alternative while awaiting the promised advanced features of the professional version.
Frequently asked questions
Question: What are the three new models recently launched by Google?
Answer: The company launched three specialized models: Gemini 3.6 Flash, Gemini 3.5 Flash Lite, and Gemini 3.5 Flash Cyber, focusing on high speed and efficiency.
Question: What distinguishes the new Gemini 3.5 Flash Lite model from others?
Answer: This model features very high speed, being able to process and generate up to 350 tokens per second, and is dedicated to massive workloads at a low cost.
Question: Why hasn’t the flagship professional version, Gemini 3.5 Pro, been launched yet?
Answer: Its launch was delayed because it failed to meet the company’s high internal goals and standards, especially regarding the efficiency and performance of writing and generating required complex code.
Question: Who are the entities that can use the new cybersecurity model?
Answer: The new cybersecurity model will be available in the initial phase exclusively and limitedly to government agencies, institutions, and trusted partners to protect infrastructure from attacks.
