Google is introducing a new AI subscription service that gives users exclusive access to its most advanced AI products.

Called Gemini Ultra, an AI-powered subscription plan priced at $249.99 per month. Currently, it is available only in the US.

This subscription provides access to new tools such as Veo 3, a video generator, the Flow video editing app, and Gemini 2.5 Pro’s Deep Think mode, which is still in testing.

Subscribers will also receive additional storage, premium access to various Google services, and enhanced AI capabilities.

🔗 Source: TechCrunch

🧠 Food for thought

1️⃣ The race for AI reasoning capabilities intensifies

Google’s Deep Think mode for Gemini 2.5 Pro represents a significant advancement in AI reasoning capabilities, matching a broader industry trend toward more sophisticated decision-making AI.

This development parallels similar reasoning capabilities being developed by competitors. MIT Technology Review highlighted AI reasoning models like OpenAI’s o1 and o3 as a key trend for 2025, noting their enhanced decision-making capabilities across various applications 1.

The industry shift toward reasoning-capable AI is driving massive infrastructure investments, with Morgan Stanley reporting that these advanced models require substantially more computational power for both training and inference 2.

According to PwC, nearly half of technology leaders now report that AI is fully integrated into their core business strategies, with reasoning capabilities becoming crucial for complex decision-making and operational efficiency 3.

This evolution marks a significant departure from earlier language models that primarily processed and generated text, moving toward systems that can analyze problems, consider multiple solutions, and present optimized answers, similar to human cognitive processes.

2️⃣ Multimodal AI emerges as the next user interface paradigm

Project Astra’s low-latency multimodal experience reflects a broader shift in how humans will interact with computers, moving beyond traditional interfaces to more natural, contextual interactions.

The integration of multiple input types (voice, vision, text) into a single AI system is gaining momentum across the industry, with multimodal models appearing in Google’s research priorities for several years 4.

This transition to multimodal interfaces is expected to fundamentally change how consumers interact with technology. Google’s preview of AR glasses with partners and the development of “Search Live,” which responds to camera inputs, highlights this trend 5.