Top News

Google AI Studio: Google has taken another big step in the race of Artificial Intelligence. Let us tell you that the company has expanded its AI portfolio and rolled out Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. Regarding which Google claims that these new models are faster, lower cost and capable of handling high-volume AI and agentic workloads better than before. The special thing is that the company has also introduced a special model Gemini 3.5 Flash Cyber ​​for cyber security.

Gemini 3.6 Flash will get better performance

It has been informed in Google’s blog post that Gemini 3.6 Flash has been prepared on the basis of Gemini 3.5 Flash. In this, coding, multimodal tasks and knowledge work have been made better than before. The company says that the new model uses fewer output tokens, making AI processing more efficient. Regarding this, Artificial Analysis Index has said that this model uses 17 percent less output tokens compared to Gemini 3.5 Flash. At the same time, on Datacurve’s DeepSWE benchmark, a reduction of up to 65 percent has been recorded in token consumption. The new model has also achieved better scores than the previous version in benchmarks like DeepSWE, MLE Bench and OSWorld-Verified.

Gemini 3.5 Flash-Lite becomes the fastest AI model

In this list, Google has described Gemini 3.5 Flash-Lite as the fastest and most affordable model of its 3.5 series. It can generate up to 350 output tokens per second. It has been designed keeping in mind agentive search, document processing and low latency AI workloads. Along with this, developers can choose different thinking levels according to their needs. The option of low level for less complex tasks and high thinking level for multi-step tasks is being given.

Special AI model launched for cyber security

Google has also introduced Gemini 3.5 Flash Cyber ​​which has been fine-tuned for identification, validation and patching of software vulnerabilities. This model will work with Google’s CodeMender Security Agent. At present, it will not be released for general users but will be made available under a limited-access pilot program for governments and trusted partners.

Preparations also started on Gemini 4 available from today

Google has said in its information that Gemini 3.6 Flash and Gemini 3.5 Flash-Lite will be available on Google AI Studio, Android Studio, Gemini API and Gemini Enterprise platforms from today. The company has also confirmed that Gemini 3.5 Pro is currently in testing with partners while the largest pre-training run for Gemini 4 has begun. But its launch date has not been revealed yet.