Another day and another advancement in the AI arms race. Google has released two new versions of its Gemini AI models, and while the announcement is mostly aimed at developers, the changes trickle down to anyone using the Gemini app or Google Search.
The main one is Gemini 3.6 Flash, which replaces the version Google put out earlier this year. The short version is that it does the same work using fewer words behind the scenes, which sounds minor but matters a lot when you’re paying by the word, as developers are. Google says it uses about 17 per cent fewer of those behind-the-scenes words than the model it replaces, and on some tasks that number climbs much higher.
For developers it’s also cheaper to run, so the apps and services built on top of it cost less to operate. It now costs $1.50 per million input tokens and $7.50 per million output tokens, which Google says brings down the overall cost of running an agentic task.
So what does this mean for you? The model is better at writing and fixing code, reading through documents, making sense of charts and data, and putting together reports. Google says it’s also gotten noticeably better at controlling a computer on your behalf, which is the work of these AI agents.
The second release, Gemini 3.5 Flash-Lite, is built for speed. It’s the fastest model Google has in this family, and it’s meant for jobs that involve churning through huge volumes of simple tasks quickly (like scanning thousands of documents or receipts). It’s cheap enough that companies can run it constantly, and Google says it’s a big step up from the previous lightweight model despite costing a fraction of what the bigger ones do.
Also announced today is Gemini 3.5 Flash Cyber. It’s designed to hunt down security holes in software and patch them, and Google is only handing it out to governments and select partners through a limited pilot, this way the bad guys don’t have access to it.
For regular users, both new models are already live in the Gemini app, with the faster Flash-Lite version also rolling out inside Google Search. Developers get them through Google’s various AI tools, and businesses can access them through Gemini Enterprise.
Google also gave a peek at what’s next, confirming that Gemini 3.5 Pro is being tested with partners now and will launch when it’s ready, while work has already started on Gemini 4.
Want to see more of our stories on Google?
P.S. Want to keep this site truly independent? Support us by buying us a beer, treating us to a coffee, or shopping through Amazon here. Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent Canadian media!
