Google’s new AI cuts costs by 600%

Hey there! Shinky here, making AI simple, useful, and powerful for you.

Google just dropped a game-changer for AI scalability, Gemini 2.5 Flash. With the clever concept of “thinking budgets,” this new model knows when to throttle its brainpower, cutting compute costs by up to 600%. It’s like your AI whispering when it doesn’t need to shout. This shift could redefine how we deploy AI in the real world which is smarter, leaner, and far more affordable.

Today’s AI Menu

• Google's Gemini 2.5 Flash Slashes AI Costs by 600% with Smart "Thinking Budgets"

• Tutorial: How to create a logo using Canva

• Everything else you should know today

• 5 new AI tools to boost your productivity

• AI Daily Prompt, Ride the Latest Social Media Trends and much more

TODAY IN WORLD OF AI

1. Google’s Gemini 2.5 flash just made AI smarter and cheaper: Big news from Google and it might quietly transform how we scale AI in the real world.

With the release of Gemini 2.5 Flash, Google has introduced a powerful new idea:
“Thinking budgets.”

What does that mean?

It’s a system where the AI intelligently dials down its reasoning power for lightweight tasks saving serious compute power in the process.

Think of it like your AI knowing when to whisper instead of shout.

Instead of brute-forcing full-blown intelligence on every task, Gemini Flash makes real-time decisions about how much reasoning is actually needed.

And the results?
A 600% reduction in compute costs for certain tasks.

Yes, 6x cheaper simply by thinking smarter, not harder.

Why This Matters?

Most of today’s AI deployments run at full throttle, even for things like summarizing an email or running a basic search. That’s like using a rocket engine to drive to the grocery store. Google’s “thinking budgets” make AI systems more modular, energy-aware, and economically scalable.

If this becomes standard practice, we could see: Massively cheaper AI for startups
▪️ Smarter agents that auto-adjust their intensity
▪️ Better personalization and energy optimization across AI ecosystems

This isn’t just a cost-saver.
It’s a mindset shift and maybe a new principle of modern AI design.

2. OpenAI’s new “Flex” processing mode is live and it’s all about patience for savings: If speed isn't essential, users can now opt for delayed responses at lower costs, ideal for background or batch AI tasks. It’s like flying economy red-eye: slower arrival, same destination, but your wallet thanks you.

3. The DOJ just dealt a historic blow to Google’s digital ad empire: A federal judge ruled that the company illegally monopolized ad tech threatening a $100 billion revenue pillar. This landmark antitrust ruling could crack open the walls around Google’s ad dominance and reshape the future of online advertising.

YOU TALK

THE AI INSTITUTE

How to create a logo using Canva

▪️ Go to Canva and sign in.

▪️ Click create a design and select Logo.

▪️ Search or browse logo templates and pick one you like.

▪️ Edit text, colors, fonts, and add icons or shapes as needed.

▪️ Align and resize elements for a clean look.

▪️ Click Download, select PNG with a transparent background, and save.

AI & TECH NEWS

Everything else you need to know today

Unfiltered: A breakthrough from DeepSeek researchers lets AI tackle previously off-limits questions. By refining data boundaries and ethical response training, models can now safely respond to sensitive topics like mental health and medical guidance without veering into danger zones.

Gifted: Google is handing out a full year of Gemini Advanced for free to every U.S. college student. This isn’t just a perk it’s a strategic leap into the hearts and minds of the next-gen workforce. Imagine having a top-tier AI co-pilot for finals, research, and side hustles.

Uplift: Hugging Face is entering the physical world by selling humanoid robots. Following a quiet acquisition, the AI company is blending its open-source smarts with real-world robotics. Think less sci-fi, more startup-ready autonomous machines built for labs, learning, and even lightweight labor.

Rewired: Microsoft just dropped BitNet B1.58, an AI model optimized to run on CPUs. That’s right, no GPUs needed. In a landscape bottlenecked by chip shortages, this 2.4B parameter model offers serious performance with old-school hardware.

Evolved: Meta’s FAIR lab just rolled out five massive human-like AI advancements. From embodied agents that learn by doing to language models tuned for emotional intelligence, this suite of releases brings AI closer to our world more aware, adaptable, and interactive than ever before.

Funding

Targeted: Theseus, a defense-tech startup, just pulled off the ultimate triple play Y Combinator, U.S. Special Forces, and $4.3M.

Insured: Vouched AI just raised $7.5M to revolutionize small business insurance.

FOOD FOR PRODUCTIVITY

5 AI Tools of the Day

🎶 Mubert: AI-generated music for content creators and streamers, providing royalty-free tracks.

📚 Wisdolia: AI-powered tool for generating flashcards from any webpage or document.

🖥️ Tome: AI-driven tool for creating beautiful presentations and story-driven content quickly.

🖼️ Fotor: AI image editing and graphic design tool with automatic enhancements and background removal.

📊 AdCreative.ai: AI for automating ad creation with data-driven design and optimization.

SPECIAL

The invisibility cloak of the AI era?

Tokenization: In a world where every byte of data is vulnerable and every AI model is listening, tokenization is stepping up as a next-gen defense layer.

Instead of encrypting data, tokenization replaces sensitive details like names, numbers, or locations with randomized, meaningless tokens. No value, no visibility, no vulnerability.

It’s like giving your data an invisibility cloak.

What makes this powerful today is its resilience against AI inference attacks and cross-platform data mining. Tokenization doesn’t just hide data, it renders it meaningless to machines, making it far more secure in hyper-connected ecosystems.

As AI scales, privacy isn’t optional, it’s architecture. And tokenization may become a core pillar of trust in every product we build.

AI PROMPT OF THE DAY

Prompt for a delivery tracking system

Prompt: 
Design a clear and efficient communication protocol for a delivery tracking system. Include steps for real-time updates, customer notifications, issue escalation, delivery confirmations, and feedback collection. Ensure the protocol covers communication between logistics teams, drivers, customer service, and end customers. Prioritize transparency, reliability, and timely updates through SMS, email, or app notifications.

If you find the contents of this email useful, subscribe now & share with your friends.

Acquire new customers and drive revenue by partnering with us

Hanoomaan is the world’s leading AI newsletter for businesses and professionals working at the world’s leading startups and enterprises. Companies like Superhuman AI, Hubspot, and The Rundown feature their products in Hanoomaan. You can learn more about partnering with us here.

Your Opinion Matters!

Got more feedback or just want to get in touch? Reply to this email and we’ll get back to you.

Thanks for reading.

Until tomorrow!

Shinky & the Hanoomaan AI team