AI Without the Cloud: Multiverse’s Tiny Models Could Change Everything

🐓🐝 AI That Fits in a Fly’s Brain? Meet Multiverse’s Tiny Powerhouses

The AI race has been dominated by bigger, faster, stronger. But now, one European startup is proving that small can be mighty.

Multiverse Computing, a fast-growing AI company based in Spain, has just launched two of the smallest yet high-performing AI models ever built — and they could change the way AI works in our daily lives.

Their names? 🐝 SuperFly (as smart as a fly’s brain, but cleverer) 🐓 ChickBrain (small enough to fit on a laptop, but with reasoning skills)

This isn’t just a quirky branding stunt. These models could mark a turning point for AI accessibility, privacy, and speed — because they can run locally on devices like smartphones, PCs, or even an Apple Watch… without an internet connection.

Why This Is a Big Deal

Most of today’s AI runs in the cloud. That means:

• You need an internet connection

• Your data often leaves your device

• Latency can be a problem

• Costs scale with usage

Multiverse’s tech flips this model. Their CompactifAI compression algorithm — inspired by quantum physics — shrinks AI models so much that they can fit on tiny chips without losing performance.

According to co-founder Román Orús, the magic is in how they approach compression:

“It’s not the typical computer science method. We come from quantum physics. It’s more subtle, more refined.”

This approach allows them to keep the accuracy and reasoning power of bigger models — sometimes even improving it — while making them small enough for edge devices.

The Two Stars of the ‘Model Zoo’

Multiverse calls their lineup the Model Zoo, with each model named after an animal’s brain size. Here are the stars:

🐝 SuperFly — 94 Million Parameters

• Compressed from Hugging Face’s SmolLM2-135

• Runs on minimal hardware (even Arduino boards)

• Perfect for simple on-device voice assistants and IoT devices

• Example: Your washing machine could understand, “Start quick wash”, troubleshoot problems, or give you usage tips — without ever talking to the cloud.

🐓 ChickBrain — 3.2 Billion Parameters

• Compressed from Meta’s Llama 3.1 8B

• Small enough to run on a MacBook with no internet

• Has reasoning abilities and performs better than the original on benchmarks like:

Imagine carrying a PhD-level assistant in your laptop — no server required.

Who’s Interested?

The list of companies talking to Multiverse is impressive:

• Apple 🍏

• Samsung 📱

• Sony 🎮

• HP 💻 (which also invested in the company)

• Toshiba 📀

These partnerships could lead to:

• Smartphones with built-in offline AI assistants

• Appliances that learn your preferences

• Wearables with instant AI feedback

• Privacy-first devices for industries like finance, healthcare, and defense

Why This Could Change the AI Game

1. Privacy & Security Running AI locally means your sensitive data — health records, personal messages, company IP — never leaves your device.

2. Speed & Reliability No internet? No problem. AI works the same in the middle of the ocean, on a flight, or in a remote village.

3. Lower Costs Reduced reliance on expensive cloud compute could make AI far more affordable for consumers and businesses.

4. Sustainability Smaller models = less energy consumption = greener AI.

The Bigger Picture: AI at the Edge

This move aligns with a growing edge AI trend — bringing intelligence to devices where the data is generated.

Industries that could benefit:

• Healthcare — Diagnosing from medical scans instantly on-site

• Automotive — Smarter, offline-capable vehicle assistants

• Manufacturing — Machines that self-monitor and troubleshoot

• Defense & Security — Field-deployed AI without cloud dependency

Challenges Ahead

While the tech is exciting, there are challenges:

• Model versatility — Small models can’t yet match GPT-5 for complex tasks

• Hardware integration — Manufacturers need to redesign for AI-optimized chips

• Ecosystem adoption — Developers must adapt to on-device AI workflows

Multiverse is betting that local-first AI will be the next big thing, and they’re positioning themselves to lead.

From Quantum Physics to Kitchen Appliances

It’s rare to see deep academic science so quickly applied to consumer tech. Orús and his co-founders — experts in quantum computing and finance — are blending theory with commercial ambition.

Their $215M funding round in June signals investor confidence, especially with backers like HP Tech Ventures and Bullhound Capital.

If their models deliver as promised, we could see AI assistants everywhere — not just in our browsers or apps, but in our fridges, coffee machines, and even light switches.

Critical Questions for the LinkedIn Community

1. Do you believe edge AI will replace cloud AI for most everyday applications?

2. How should companies balance the trade-off between model size and intelligence?

3. What new business models could emerge if AI runs entirely on personal devices?

4. Could offline AI solve some of the privacy issues that regulators are worried about?

5. Which industries will benefit the most from this kind of AI miniaturization?

AI has been chasing bigger and better. Maybe the future will be smaller and smarter.

What’s your take — is this the beginning of the “Small AI Revolution”? 🚀

Join me and my incredible LinkedIn friends as we embark on a journey of innovation, AI, and EA, always keeping climate action at the forefront of our minds. 🌐 Follow me for more exciting updates https://lnkd.in/epE3SCni

#AI #ArtificialIntelligence #EdgeAI #IoT #MachineLearning #Privacy #DataSecurity #QuantumComputing #OnDeviceAI #TechInnovation #AITrends #SmartDevices #AIInBusiness #StartupFunding #AIModels #ModelCompression #CloudComputing #AITransformation #AIRevolution #AIPrivacy #SmallAI #AIHardware #Innovation #DigitalTransformation #TechnologyTrends #AI2025

Reference: Tech Crunch