Private AI · No Cloud · Physical Device

ANV-GYRO YOUR AI. YOUR DESK. YOUR DATA.

Runs Ministral 3B entirely on-device.
No data transfer to server. No subscriptions.
Embedded AI in physical hardware — made in the EU.

Ministral 3B On-device inference No internet required 🇪🇺 Made in EU 🇪🇺 Hosted in EU 3× Raspberry Pi 5
How it works Request a demo
↓ explore
No Data Transfer to Server Ministral 3B On-Device Made in EU Hosted in EU Embedded AI Hardware GDPR Compliant by Design Sovereign AI No Data Transfer to Server Ministral 3B On-Device Made in EU Hosted in EU Embedded AI Hardware GDPR Compliant by Design Sovereign AI
// 01 — Privacy Architecture

Zero data
leaves
your desk.

Most AI sends every query to servers you don't control. Every document you upload, every question you ask — it travels the internet.

ANV-Gyro is different. The model runs entirely inside the box on your desk. Nothing leaves. Ever.

🔒

No data transfer to server

Inference runs on-device. Your queries, documents, and outputs never touch an external server.

📡

Works fully offline

No API calls, no cloud dependency. The model lives in flash storage — available 24/7, no internet required.

🏛️

GDPR compliant by design

Data residency is physical — your building, your device. Compliance is structural, not contractual.

No rate limits, ever

No monthly quotas, no usage billing, no outages. Run inference as hard as you want, all day.

RPi RPi RPi ANV-Gyro Ministral 3B · on-device Data stays here. You Cloud 3rd Party no cloud no 3rd party
// 02 — AI Engine

The model
that fits
on your desk.

Ministral 3B
by Mistral AI · Paris, France 🇫🇷

A 3-billion parameter language model optimised for edge deployment. Fast enough for real conversation, small enough to run on commodity hardware — without sacrificing intelligence.

3BParameters
~2GBModel size
<1sFirst token
EUOrigin

Runs on the 3× RPi 5 cluster via quantised inference (GGUF / llama.cpp). No GPU. No API key. No billing. The model lives in flash storage — yours to use indefinitely, fully offline.

Use cases
Document summarisation
Code assistance
Local Q&A assistant
Data analysis
Private note intelligence
Who needs this
Legal & finance firms
Healthcare providers
Government agencies
R&D departments
Privacy-first individuals
// 03 — Origin & Hosting

Made in EU.
Hosted in EU.

From hardware assembly to infrastructure, ANV-Gyro never leaves European jurisdiction. Built under EU law. Backed by EU values.

★★★★★
★★★★★
★★
🏭

Manufactured in EU

Hardware designed, assembled and tested within the European Union. Full supply chain transparency.

🖥️

Hosted in EU

All cloud services (firmware updates, optional sync) run on EU-based infrastructure. No US cloud providers.

📜

GDPR by design

No personal data on external servers. Privacy is an architectural property, not a policy document.

🤖

EU AI Act ready

Transparent model provenance (Mistral AI, FR). On-device deployment minimises regulatory exposure.

🔬

European AI model

Ministral 3B is by Mistral AI — Paris-based, Europe's leading AI lab and answer to Big Tech AI.

🛡️

Sovereign AI

No dependency on US or Chinese AI infrastructure. Your organisation's intelligence stays sovereign.

// 04 — Engineering View

See every
component.

Interactive engineering diagram. Toggle the display, projector, and load simulation to see how ANV-Gyro behaves.

Projection RPi 1 RPi 2 RPi 3 IN OUT Flip-up display (10") Hinged · 70° viewing angle Sphere L (intake) Spins with CPU load Sphere R (exhaust) Glossy black · spins Pico projector DLP2000 Camera OV5647 · magnetic Microphone MEMS · pogo-pin Base (blue acrylic) 300×240×80 mm · 3× RPi 5 Round displays (2×) GC9A01 · 1.28" ⚡ Heat → fans → spheres spin = physical load indicator Ministral 3B · on-device · no cloud · ANV-Gyro
// 05 — The Hardware

Built to last.
Built to see.

Sphere rotation speed is a direct physical readout of AI workload. No dashboard needed — visible from across the room.

01

Airflow-driven spheres

Two Ø80 mm piano-black aluminium spheres. Turbine vanes convert fan airflow to rotation. Idle: 2 RPM. Full AI load: 25 RPM.

Pure physics
02

Flip-up 10" display

IPS panel hinged at rear edge. Closed = silent blue slab. Open = full workstation at 70°. Friction-lock hinges.

1920×1200
03

Blue acrylic housing

Translucent sapphire-blue PMMA, 300×240×80 mm. RGB underglow. RPi boards visible through the material.

PMMA · EU made
04

3× Raspberry Pi 5

Cluster compute for distributed inference via llama.cpp. PWM fan control from GPIO temperature sensors.

4GB each
05

Round status displays

Two GC9A01 1.28" circular displays in the front face. CPU load and clock always visible.

GC9A01 · 240×240
06

Pico projector

DLP2000 rear-mounted, projects upward onto the wall behind the device. Present AI outputs without a second screen.

DLP2000
// 06 — How It Works

Four steps.
Zero cloud.

1

You type a query

Via the flip-up display terminal, local network web UI, or REST API. No account. No cloud login.

2

Ministral 3B runs locally

The model processes your request on the 3× RPi 5 cluster. Quantised inference via llama.cpp. No internet needed.

3

Spheres spin with load

Fans speed up with temperature. Airflow spins the turbine-vaned spheres. You see the AI working — physically.

4

Answer appears. Data stayed home.

Response is generated and shown. Nothing left the device. No server log. No trace. Spheres slow down.

// Get in touch

Your AI.
Your desk.
Your data.

We are looking for pilot customers, enterprise partners and early adopters who believe private AI should not mean slow or compromised AI.

Contact us →
🇪🇺 Made & hosted in EU · 🔒 No data leaves your device · ⚡ Ministral 3B on-board