About the Archive

Then the pandemic hit and the field was off limits. We were stuck indoors. The conversations moved online and went deeper than they ever had in person. Somewhere in those long months we all kept running into the same wall, the same missing thing, and that absence became the seed of everything you see here.

What we wanted was simple to describe and impossible to find. One place that puts the entire military inventory of the world under a single roof. Every weapon, every country, every defense company, every conflict, all held in the same structured data, all of it laid side by side so you can actually compare them. A resource that scores each system on one consistent scale and runs the price and performance through real calculation and proper algorithms, then hands you the result in plain sight.

It was not out there. The knowledge had been scattered to the four winds. Some of it sat locked behind costly corporate databases. Some had been left to rot on websites that stopped being updated a decade ago. The rest was sprinkled across hundreds of sources that had no idea the others existed. So we stopped waiting for someone else to build it and decided to build it ourselves.

The first run was small and it ran into a wall fast. A few of the people we started with drifted away as life pulled them elsewhere. The developer we had on board back then was simply not equal to a system of this size. And the deeper lesson landed hardest of all. Gathering and processing this much data by hand was flatly beyond what human hands can do.

The second run was wiser, and that was when the true scale of the thing finally showed its face. We were determined that every record we wrote would be its own creation. We wanted something synthesized, cross referenced, and put into our own words from the ground up. To get there we would have to sweep through every printed and written source across a century and a half and turn all of it into something a database could actually hold. Standing in front of that mountain, with the tools we had at the time, we admitted it was bigger than us.

Both attempts pointed to the same answer. Patience. We put our faith in time, and in a simple bet that artificial intelligence would grow far stronger in the years ahead. We shelved the project on purpose and waited for the moment to be right.

That moment arrived. The AI models caught up to the work we had been dreaming about three years earlier. In January 2026 we started gathering data in earnest. The archive in front of you is what came of that decision.

There is a serious stack of technology behind us and an operation that runs around the clock. The whole thing rests on one clean line drawn in the sand. Artificial intelligence does the heavy lifting of gathering, scanning, and processing the data. The writing in the encyclopedia, the judgment calls, and the final say belong to us, to human hands.

We work our way, methodically, through nearly every printed book, magazine, and newspaper put out over the last 170 years. We reach into the spoken memory of the internet too. We are constantly combing YouTube, forums, specialist sites, and Twitter for anything worth keeping. To us YouTube is the largest oral archive of military history anywhere on earth. We pull the captions from tens of thousands of documentaries, interviews, and firsthand accounts from people who were actually there, turn each one into a document, and fold them slowly into our verification system. Because the real range of a tank sometimes never makes it into a single book. Sometimes it lives in one offhand sentence from a man telling a camera about the machine he used to drive.

Right now we are sitting on hundreds of terabytes of raw material. To get through it we run the most capable AI models in the world, around the clock, every day of the week, and we spend real money doing it. While we sleep, the system keeps reading. And this is not one model wearing every hat. Seventeen separate AI agents work side by side, each one trained on its own narrow job. One hunts down sources. One pulls technical specifications. One stitches together the web of relationships between a war, a weapon, a country, and a company. One tracks down imagery. One sits over all of it, checking the rest against itself.

We pick the right model for each job. Reasoning, verification, and the careful text work go to the Claude family. Wide context reading and extraction go to Google Gemini. Fast structured extraction goes to Mistral. The high volume bulk sweeps run on Llama through Groq and Cerebras. Multilingual work and translation go to Qwen. Independent cross checking goes to GLM. The best models in the business, each one set loose on the task it does best.

The system does far more than hoard. It argues with itself all day long. When three sources cannot agree on the weight of a single weapon, it raises a flag. A verification agent reasons its way toward whichever figure stands on firmer ground. And the last word, always, comes back to us.

There is a long road behind every weapon on this site. It starts as fragments pulled from dozens of sources. It goes through cross checking. Its specifications get pinned to a common scale. It gets scored. It gets wired to the wars it fought in, the countries that fielded it, and the company that built it. Then a person sits down and writes it, and it gets carried into more than one language. Finally it goes live, and even then we never call it finished. Every record is a living draft, and it keeps on growing.

Here is a small confession that tells you everything about how hard this work is. We spent a solid month arguing among ourselves about nothing but the categories. Military history refuses to sit in tidy boxes. Does a self propelled gun count as artillery or as armor? Is a battlecruiser a cruiser or a battleship? We are still hashing out which weapon belongs in which bracket. The question is open, and it will stay open for a long while yet.

An absolute approach to scoring sits at the core of the archive. Most sources rate a weapon through the eyes of its own era and leave it there. We went the other way. A Mark I built in 1917 and a T-14 Armata built in 2026 get weighed on the very same scale from 0 to 100. That is how systems from every corner of history finally become things you can hold up against one another.

We run the price and performance through serious calculation and real algorithms as well. The true worth, though, lives in the threads between the records. This weapon fought in that war. That country was part of it. That country bought from this company. This company has that rival. We spend our days weaving the connective tissue of military history. And we build every record from nothing, in language that is ours.

Our team is solid, and it is still small. We will say that to your face. The scale of this thing reaches miles past what any one person could ever hold in their head. Nobody alive knows land, sea, air, cyber, and space all at once. Now stack on top of that the history, the economics, and 170 years of military development across 190 countries. That is exactly why we lean on a team of many trades and on the muscle of the machine at the same time. It is the only way a handful of obsessives ever produces the work of hundreds.

We are from Turkey, and we are proud to say it. Keeping the archive global and even handed is the principle we will not bend. We catalog even the nations Turkey is at odds with, and we hold them to the very same standard. A weapon or a country earning a place in this archive is a statement of record. It is information, plain and simple.

Let us be plain about this. We never take the side of war. We catalog these machines and their history to understand them and to set them down on the record. Studying the armor on a tank says nothing about wanting one. Knowing history is a world apart from wishing to live through it again. Our aim is to understand, to document, and to hand accurate knowledge down to whoever comes after us.

Being straight with you matters to us, so here it is. This site is still under construction. We had actually penciled in 2028 as the year to open the doors, by which point the content would have ripened. Then the research we did and the counsel we got pointed us somewhere else. Opening early and building our visibility, our search ranking, and our traffic a little at a time made far more sense than waiting for perfect behind a locked door. So here we are. Half built and wide open. We are filling our gaps right out in the open, in full view, with nothing tucked away in the back room.

The target is clear. By 2030 we mean to have the archive roughly 90 percent complete and running at 99.99 percent accuracy. Between now and then we will keep grinding through those hundreds of terabytes, correcting as we go, day after day, and deepening what is already there. Even so, an honest word from us to you. This archive will most likely never be truly finished, and there may always be a few errors hiding in it. We treat that as the nature of a living work.

We move slowly because we mean to get it right. Rush it and all you produce is one more forgettable copy. We are building something that lasts.

At the moment there is no outside money at all. Every cost comes straight out of our own pockets. We reckon the yearly running cost at somewhere near the price of a car. Even so, there will be no Google ads on these pages. We have no intention of turning the archive into a cheap billboard.

To keep this thing alive over the long haul, we are open to a limited form of sponsorship. We would only ever accept it from companies that fit the subject of the site and already sit in our database. And there is one red line here that does not move. No sponsorship touches the scoring, the rankings, or the price and performance verdict. Every sponsorship is disclosed out in the open. A company paying us never shifts its score by so much as a single point. No government, no army, and no manufacturer bankrolls us. Our independence is the ground our neutrality stands on.

The team has gaps, and we need good people to fill them. If you are reading this with a real fire in you for OSINT, for war history, or for the machines themselves, and you want a hand in something like this, get in touch. Convince us of what you can do and we would be glad to have you with us.

We especially need help carrying the archive into new languages. We are after people who have a real command of military terminology and speak the target language as natives. Where it makes sense, we are happy to put that work on a paid footing. For now we are live in Turkish and English. In time, we mean to open the doors to the other languages of the world.

Our ambitions run well past this website. One day we want to turn the archive into a printed military encyclopedia, volume after volume. Further down the line, a documentary series and a YouTube channel are not off the table. For now those are distant lights on the horizon. We think in the long term, and we mean to carry this as far as it can possibly go.

Here is our word to you. Every record comes with its source. Every error gets fixed the moment we catch it or you point it out. No country and no flag gets the benefit of the doubt over another. The data stays independent, full stop.

And here is what is still coming. Full coverage. A clean sheet with no errors. Speed. All of it ripens with time, and we are counting on your patience while it does.

What you find here is for reference and for learning. Please treat it that way and keep it clear of any professional or operational decision.

Spot an error and tell us. The readers who catch the most mistakes turn out to be the best company we have on this road. This whole thing is the child of an obsession. What we want, all of us together, is to leave behind a work we can be proud of for the rest of our lives. Thank you for walking part of it with us.