Google appears to be working on an upgraded version of Bard called "BardAdvanced," developer Dylan Roussel shared the news on X (formerly Twitter). According to Roussel, BardAdvanced appears to be powered by GeminiUltra, the highest level of Google's latest release of large language models (LLMs).
This is an upgrade from the current version of Bard, which runs the mid-tier Gemini Pro. A screenshot shared by Roussel describes BardAdvanced as "a more powerful large language model with advanced math and reasoning capabilities." Developer Bedros Pamboukian also discovered a string of code indicating that Google is developing a "premium" package.
In addition, Russell spotted more updates and shared screenshots, including a feature codenamed "Motoko" that allows users to create custom AI bots. However, there's no indication yet if users will be able to share these bots, or if they'll need to pay to use the feature. Last year, OpenAI announced it would allow GPTPlus users to create and share custom robots.
Google may also launch a new "upgrade" feature that uses artificial intelligence to expand and improve the prompts users send to Bard. As Russell discovered, Google also mentions a "Gallery" section, which he speculates will let you "explore different themes and see what you can do with Bard."
Other small tweaks revealed in the screenshot include a new "Tasks" tab, which may be used to track longer-term work that Bard is currently working on, such as image generation, and a way to add custom backgrounds to shared Bard prompts.
We don't know yet when (or if) these features will become official. Over the past few months, Google has made a series of upgrades to Bard, such as allowing it to respond to queries and answer questions about YouTube videos in real time, as it looks to keep up with the likes of ChatGPT and Microsoft Copilot.