Gemini's paid subscribers can now try an experimental Gemini 2.0 Advanced model

It’s been a busy week over at Google. Not too long after launching Gemini 2.0 Flash (Experimental), the Mountain View tech giant is now opening up Gemini-Exp-1206, Gemini 2.0 Advanced’s experimental model, to all Gemini paid subscribers.

Google says that Gemini-Exp-1206 excels in complex tasks like coding, math, reasoning, and instruction. It can also help users create business plans and solve problems in general.

But, as an experimental model, you may still experience a few hiccups here and there, like the lack of real-time information access and incompatibility with some users. If you’re a Gemini Advanced paid subscriber, you can take this model out for a spin via the Gemini dropdown on desktop and mobile web.

Gemini 2.0 Flash, on the other hand, is available to all Gemini users, including those in the free tier. It’s a chat-optimized version of Gemini 2.0 that offers speed, which is different than the Gemini-Exp-1206 that more serves as a big jump in performance in complex tasks.

Last week, Google also launched Deep Research, an AI assistant for Gemini Advanced subscribers, which does your research for you. It generates comprehensive, multi-step research reports by analyzing web data.

That, is thanks to advanced reasoning and a 1 million token context window to create these detailed reports with links to original sources.

The company also launched Veo 2, a video-generating model, amid hefty competition from OpenAI and its hotly-hyped Sora launch. But, unlike the first Veo model, Veo 2 can create AI videos up to 4K in resolution.