Outside
A small, familiar object. It sits in a pencil case without announcing itself and does not read as a gadget.
In development - early access open
AI built into everyday objects.
CloakMind Tape puts a camera, a touchscreen and multimodal AI inside something you would never look at twice. Point it at an exercise and the worked answer appears on the device.
Early access locks the 99,90 € founder price
Free to join - no payment now - one email when orders open · Privacy
Illustration of the flow on the device. Firmware in development.
The reveal
The whole idea of CloakMind is the gap between what a device looks like and what it does.
Outside
A small, familiar object. It sits in a pencil case without announcing itself and does not read as a gadget.
Inside
A 5 MP autofocus camera, a 2 inch touchscreen, Wi-Fi and a connection to multimodal AI that reads whatever you point it at.
Working prototype
The firmware, the camera, the touchscreen and the service that reads pages all run today. What is left is the enclosure, industrial design and a production run. Real hardware photos go here as they are ready - we are not putting up renders that pretend to be a finished product.
Current prototype hardware
How it works
01
Point CloakMind Tape at the exercise and tap. A live viewfinder on the screen lets you frame it properly.
02
The photo goes over Wi-Fi to a multimodal AI that reads the page and works out what it is looking at.
03
The worked answer appears on the device screen - method, key steps, result. Around 30 seconds on the current prototype.
04
Ask a follow-up on the touchscreen. It keeps the context of what you captured, so you don't start over.
Illustration of the flow on the device. Firmware in development.
Discretion
CloakMind Tape looks and feels like an everyday object while putting powerful AI directly in your hand.
Compact, familiar and quiet. The technology is built into the object instead of announcing itself.
No notifications, no feed, no apps. One screen that does one thing, so you stay on the page in front of you.
Camera, screen and controls are all on the device. You never unlock a phone to look something up.
CloakMind is built for study, revision and settings where AI assistance, reference material or external tools are permitted. Always follow the rules your course or institution sets.
Use cases
Multimodal AI works out what it is looking at, so you don't pick a mode first. Illustrative examples of the kind of work it is designed for.
MathsSolve the integral of x ln(x) dx
Integration by parts with u = ln(x): the result is (x^2/2) ln(x) - x^2/4 + C, with each step shown.
PhysicsA 2 kg block on a 30 degree incline. Find the acceleration.
Free body diagram, forces resolved along the plane, then a = g sin(30) with the numbers substituted.
ProgrammingWhy does this loop never terminate?
Reads the code off the page, points at the counter that is never incremented and shows the fix.
LanguagesRewrite this sentence in the passive voice
The rewritten sentence plus the grammar rule behind it, at the level you work at.
Multiple choiceWhich of these four options is correct?
The option it picks and, more usefully, why the other three are wrong.
Illustrative examples of supported work, not screenshots of the final interface.
Follow-up
Most tools stop at photo in, answer out. CloakMind keeps the capture in context, so the second question still makes sense.
You scroll and navigate the answer on the touchscreen, so a long explanation is not a problem.
Initial capture
Solve this integral.
Follow-up
Why did you substitute x here?
Why dedicated hardware
Your phone can do this. It just makes you do six things first, and hands you a feed on the way.
Pick it up and the viewfinder is live. No unlock, no app, no permission prompt.
There is nothing else on the device to pull you away. That is the point of it not being a phone.
Autofocus and image processing tuned for printed exercises, not for faces or landscapes.
Step by step by default, at the depth you set - not just a final number.
How it compares
| How it compares | Smartphone | AI calculator | CloakMind Tape |
|---|---|---|---|
| Camera capture | Yes, after unlocking | Usually not | Yes, ready |
| Manual typing | Often | Yes | No |
| Follow-up with context | Depends on the app | Limited | Yes |
| Distractions | The whole phone | Few | None |
General comparison of device categories, not a claim about any specific product.
The AI
Subscribe to CloakMind AI, or bring your own API key and pay us no subscription at all. The hardware works the same either way.
0 €per month
Use your own compatible API key and your own provider. No subscription to us.
For technical users
4,99 €per month
CloakMind AI with a monthly allowance of sessions and model routing handled for you.
First month included with the device
9,99 €per month
A larger monthly allowance for heavier use, with the same routing and follow-up context.
For frequent use
What counts as a session
One photo, the answer it returns, and follow-up questions about that same capture.
Plan prices are planned, not final. They will be confirmed before launch, and early access members hear first.
Early access
Joining takes one field and about three seconds. No payment now.
Founder early access
99,90 €
119,90 € Planned retail
The first production run is small because we assemble it ourselves, so founder pricing covers that first run.
Free to join - no payment now - one email when orders open · Privacy
No payment, no card, no commitment. You're telling us you're interested, that's all.
Questions
Yes. The AI runs in the cloud, not on the device, so it connects over Wi-Fi - typically your phone hotspot or any available 2.4 GHz network. Reference notes you have saved to the device can be read with no connection, but solving an exercise needs one.
It is designed to work with multimodal AI across maths, physics, chemistry, programming, languages, multiple choice and general printed text, from secondary school through university level. The model works out what it is looking at rather than making you pick a mode first.
The current prototype typically returns an answer in around 30 seconds. We are not going to quote a number we cannot hit today, and we will update this as the firmware and model routing improve.
Yes. CloakMind keeps the context of your original capture, so you can ask why a step was taken instead of starting again from a new photo. You scroll and navigate the answer on the touchscreen.
Yes. Technical users will be able to bring their own compatible API key and provider, with no subscription to us. The device works the same way either route.
No. You can use CloakMind AI or configure your own API key. The first month of CloakMind AI is included with the device. Plan prices are planned rather than final and will be confirmed before launch.
There is no launch date yet and we are not going to invent one. The planned price is 119,90 €, with 99,90 € for founder early access. Join early access to be notified before the first release.
Only where the rules of your course or institution allow external tools or AI assistance. Plenty of assessments expressly permit reference material or AI help - that is what CloakMind is built for. Using it where it is not permitted breaks those rules, and we do not design or market it for that.
One email field. No payment. Lock the 99,90 € founder price before the first run.
Free to join - no payment now - one email when orders open · Privacy