OpenAI recently launched ChatGPT Agents, and now it seems like its biggest rival, Google, is also working on the same. According to insights shared by TestingCatalog, Gemini’s Agent Mode was initially planned as an early-access feature for Ultra-tier subscribers. However, with OpenAI’s Agent Mode already making headlines, Google may be forced to extend access to Pro users as well.
If we talk about Gemini’s upcoming Agent Mode, then it will be much like ChatGPT’s, and it is expected to go beyond simple Q&A interactions. While full details are still under wraps, it is expected to be designed for executing multi-step tasks, suggesting a direction similar to OpenAI’s approach, automating workflows like browsing, file management, or code execution.
BREAKING 🚨: Google is preparing to release Deep Think on Gemini in the coming weeks and working on a new Agent Mode!
— TestingCatalog News 🗞 (@testingcatalog) July 10, 2025
Deep Think on Gemini performs very close to the leaked "Kingfall" model.
What's Agent Mode? Check below 👀 pic.twitter.com/kuppTgkjXA
Another tweet from a week earlier also confirmed that Google is working on this new Agent Mode alongside something called “Deep Think,” a tool reportedly close in performance to the previously leaked “Kingfall” model.
OpenAI’s Agent Mode allows ChatGPT to perform end-to-end tasks such as booking appointments, creating editable documents, or running code on a virtual system. With this move, Google is clearly gearing up to offer a similarly autonomous experience inside Gemini.
While no public launch timeline has been shared by the company, the competitive landscape indicates that the Agent Mode on Gemini could roll out sooner rather than later.
Apart from the Agent Mode, Google has been constantly releasing updates to strengthen Gemini’s overall capabilities. The tech giant recently rolled out new AI-powered video creation tools within Gemini and Flow. These tools allow users to turn a single static image into an eight-second video with sound using a simple prompt. Currently available to Pro and Ultra subscribers in select countries, the feature builds on Veo 3, Google’s most advanced video generation model to date.
According to the company, more than 40 million videos have already been created across Gemini and Flow in just seven weeks of its launch. And now the tool is being used for everything from animating drawings to generating ASMR-style scenes.
That’s a wrap about Gemini for now. But if you want the next update faster, then join us on WhatsApp. We share all the latest tech and AI news along with the in-depth reviews, analysis, and more.
