Japanese Armoured car Japan (1941-1944)
Super Heavy Tanks – 1 built


The tanks talked about in this article are very obscure, with very little being written on them and almost no original documents surviving. The information presented here has been cobbled together to the best possible example with what survives on these tanks. This was not aided by the mountains of misinformation that have sprouted in recent years from rekindled interest in these tanks. Information for the Mi-To leans heavily on Shigeo Otaka, as he is the only primary source, having been there. However, it must be said that the recollections of an engineer do not mandatorily reflect the historical truth, as they might be distorted by time or other things.

Talking about these vehicles is further complicated by the fact they are often misnamed. Designations such as O-I Experimental, O-Ni and O-Ho have appeared, which have no historical backing and can confuse any reader that tries to use different sources.

Go Big or Go Home

After Japan’s defeat at the Battle of Khalkhin Gol, the army realized its current tanks, the Type 97 Chi-Ha and Type 95 Ha-Go, were no longer competitive against the more modern tanks of the Red Army. Hideo Iwakuro, Chief of Army Affairs, instructed Colonel Murata, head of the 4th Technical Research Group, to construct a super-heavy tank. Hideo Iwakuro’s exact words were “I want a huge tank built which can be used as a mobile pillbox in the wide open plains of Manchuria; Top secret.” “Make the dimensions twice that of today’s tanks”. At the time of this request, the biggest tank Japan had built was the modest Type 95 Heavy Tank. This leap in size from 26 to what would be 100 tons did worry Colonel Murata’s engineers, but each man kept their doubts to themselves.

Please note that most people know this tank as “O-I”, indeed it may have been referred to as such historically, but this has not been proven. For clarity here it is referred to here only as “Mi-To”.

The Mi-To Super Heavy Tank

In March 1941, the initial design work was complete and the tank was ready to be built. The following month, April, several select engineers, including Shigeo Otaka, the main source for the development of the Mi-To, were taken to the 4th Technical Research Group’s Headquarters in Tokyo.

They were told not to speak of what they would see inside, as it was top secret. The construction engineers were guided through the cramped and dimly-lit barracks where they would later conduct meetings on the tank’s construction. At the back of the barracks, they were taken into a soundproof room with no windows.
The only entry to the room was a set of double doors (like an airlock, where you go in one door, close it, then go in the second door), designed to keep people from seeing in. Each officer present at the meeting had a piece of the design, that after some effort to assemble, revealed the tank. The tank’s name was Mi-To, for Mitsubishi, the company behind the construction and design, and the city, Tokyo.

Engineer Shigeo Otaka gave the tank’s parameters as 10 meters in length, 4.2 meters in width, and 4 meters in height (or 2.5 meters without the turret). The width of the tracks was to be 900 mm. Propulsion would be provided by two Kawasaki Ha-9 air-cooled gasoline engines, placed parallel to each other lengthwise in the hull, each producing 550 hp for a total of 1,100 hp.

The transmission was a scaled-up version of that used on the Type 97 Chi-Ha. It was mounted in the rear, between and behind the engines; it had 5 forward gears plus reverse. The suspension consisted of two coil spring bogies per side, each having two sets of four all-steel roadwheels (four on the inside of the track teeth, four on the outside) for a total of 8 pairs of roadwheels per side.

The tank was armed with a Type 96 15cm howitzer in the main turret. In front of the main turret were two mini-turrets, each having a 47mm Type 1 Tank Gun. A fourth turret with dual Type 97 machine guns was placed above the transmission to the rear.

The armor was appropriate for a mobile bunker; 150 mm at the front, made by bolting an additional 75 mm plate to the tank’s 75 mm frontal hull. The side hull armor was only 35 mm thick, with a 35 mm sideskirt covering the entire side and tracks.

On the inside, there was enough room for a man to stand comfortably. Two 16 mm bulkheads (other sources claim 20 mm) divided the tank into three sections, driver’s compartment, main turret fighting compartment, and engine compartment.


The following dates are unsourced. On April 14th, 1941, construction started on the Mi-To; Colonel Murata’s plan was for construction to last only 3 months. Problems were found with the tank’s cooling system which delayed construction until January 1942. The hull was completed on January 8th. Meanwhile, Mitsubishi made the turrets.

For testing, the Mi-To was shipped to Sagami Armory in Sagamihara, 51 km (31.7 miles) south of Tokyo. Only the people involved in the development of the tank participated in the transport, making it very difficult. In June 1943, the tank was disassembled for transport and covered with awning to keep it hidden. Work on moving the tank started in at 2:00 am every day and lasted until dawn; this took 10 days. It arrived at Sagami Armory at the end of June, and tests were scheduled for August 1st.


The Mi-To Super Heavy Tank’s trials took place without the additional 75 mm armor and without the main turret, as the 35 mm roof plate was not ready. These exclusions meant the tank weighed only 96 tons. On the day of testing, Colonel Murata was out at the front, so in his place for observation was Lieutenant Colonel Hidemitsu Nakano. Also present was the chief of Sagami Armory was, Tomio Hara.

The testing ground was quite soft so during off-road portions of the test the tracks sunk into the ground up to a meter deep. The tank tried to wiggle itself free but this just caused it to sink further, as well as cause damage to the suspension. After this test, a full bow with both hands to the ground was made to the repair department chief.

After the tank was pulled out, tests continued on concrete road surfaces not soft ground. Because of the damage to the suspension, the tank only tore up the concrete and damaged its suspension further. Tests were abandoned. The tank was covered with a tarpaulin and left alone. It was finally scrapped.

The Mi-To was disassembled for scrap either between August 3rd and 8th of 1943, or more likely, in 1944. All that remains of the Mi-To is a single track link, measuring about 800 mm in width, and 300 mm in pitch. It was previously at Wakajishi Shrine but has since been moved to JGSDF Camp Takigahara.


There was a post-war questioning with four Mitsubishi engineers who had worked on the Mi-To, the only one who has been identified so far is Shinjo Masahisa. Unfortunately, these men did not have much to say about the Mi-To, and what they do say conflicts with proper sources. The only information they could provide was that the Mi-To had 100 mm of frontal armor, and could reach 40 km/h on roads. Both of these numbers seem incorrect as all other sources state the tank’s armor was 75+75mm and it had a top speed of 40 km/h; such speed seems very optimistic for such a heavy tank.

Tomio Hara reports the Mi-To as having a 10 cm (10 cm often refers to 105 mm in Japanese gun terminology) main gun; but blueprints for the tank show a short 150 mm howitzer. There is also an (unsourced, again) size claim for a length of 10.1 meters, width of 4.8 meters, and height of 3.6 meters.

Mi-To specifications

Dimensions (L x W x H) 10 x 4.2 x 4 m (32.8 x 13.8 x 13.1 ft)
Total weight, battle ready Over 100 tons
Crew 11 (driver, co-driver, commander, main turret gunner, 2x main turret loaders,
2x 47 mm turret operators, machine gun turret operator, radioman, engineer)
Propulsion 2x 550 hp Kawasaki Ha-9 air-cooled gasoline engines
Suspension Coil springs
Speed (road) 30 km/h (18.6 mph) (Probably an optimistic number)
Armament 150 mm (5.9 in) Type 96 howitzer
2x 47 mm (1.85 in) Type 1 tank guns
2x 7.7 mm (0.3 in) Type 97 heavy tank machine guns
Armor 75+75 mm (2.95 + 2.96 in) frontal armor
35+35 mm (1.38+1.38 in) side armor
Total production 1 incomplete prototype

O-I: Monster in the Mist

By 1944, the situation for Japan was desperate; they had woken the sleeping giant, the US, and were steadily losing occupied islands. News had come from Germany about a super-heavy tank and this rekindled the Japanese Army’s interest in the idea. Development then started on a new super-heavy tank, the O-I. The name O-I comes from the Japanese tank naming nomenclature, O meaning “heavy” and I meaning “number one”, or “first.” These tanks have been called “O-Ni” (the 100 ton), and “O-Ho” (the 120 ton), but these designations are completely fictional, and were never used for these tanks.

There are several different designs for the O-I. The first is a 100-ton tank. There is very very little about the 100 ton O-I, so little that it is possible that it never existed at all. The accompanying diagram and artist’s interpretations based on the diagram are solely representative of this tank. It is also possible that this design was a concept for the Mi-To; or that it is just fiction. The reason it is included here, under the O-I designation, is for the gun; it appears to be armed with a Type 92 10 cm cannon.

The real O-I is the 120-ton version. Tomio Hara is the primary source for this tank; he reports it as being armed with a Type 92 10 cm cannon. Secondary armament consisted of three mini-turrets. The two forward mini-turrets were offset to the left; so that the midpoint of them was slightly left of the centerline of the tank. One was armed with a 47 mm Type 1 Tank Gun, the other was armed with a Type 97 machine gun. The third mini-turret was in the rear, armed with dual Type 97 machine guns. Armor was up to 200 mm thick. The powerplant was the same as the Mi-To, two 550 hp Kawasaki Ha-9 engines. Top speed was to be an again optimistic 25 km/h. The tank would have had a length of 10 meters (another source says 11 meters), width of 4 meters, and height of 4.2 meters. Track width was to be 750 mm, riding on a suspension made of coil springs. Had the O-I been built, it would have been made by Mitsubishi; however due to the war conditions, the project was canceled.

O-I specifications

Dimensions (L x W x H) 10 x 4 x 4.2 m (32.8 x 13.1 x 13.8 ft)
Total weight, battle ready Over 120 tons
Propulsion 2x 550 hp Kawasaki Ha-9 air-cooled gasoline engines
Suspension Coil springs
Speed (road) 25 km/h (15.5 mph) (Probably an optimistic number)
Armament 105 mm (4.13 in) Type 92 cannon
47 mm (1.85 in) Type 1 tank gun
3x 7.7 mm (0.3 in) Type 97 heavy tank machine guns
Armor Up to 200mm
Total production None

Type 2605: Forgotten Goliath

These two designs recently came out of the Russian archives, from a single post-war report by a Soviet engineer named Grigoriev. The report is mostly about the cooling system used on the Type 2604 and Type 2605 tanks. These tanks have also been called Type 4 and Type 5; these names are simplifications of the complicated “Type” naming system that the Japanese used. The Type 2604 was powered by two V12 BMW marine engines, each having three radiators plus an oil radiator. The engines were cooled by two central ventilators, powered by the drive train. Air was sucked in through large grilles on the engine deck.

The Type 2605 had a 30% smaller engine compartment volume while keeping the same engines and cooling abilities. The cooling system ventilator was 950 mm in diameter. There was a wooden mockup built that replaced the two BMW engines with a single Daimler-Benz V12 water-cooled diesel engine. Ventilation of the crew compartment was good, but only while the engine was running. The disadvantage of this system is that to service the engine the radiator had to be removed.

There is no other information about these tanks; the only sources to attest to the fact they existed is the aforementioned Soviet document and the following photograph. The design of this bunker turret at an unknown location in Manchuria is strikingly similar to that of the Type 2604/5.

If this photograph is real, it is possible that this was a Type 2604 or 2605 turret used as a bunker.
If this photograph is real, it is possible that this was a Type 2604 or 2605 turret used as a bunker – Source


Tank and Tank Battles – Shigeo Otaka
Pacific War Secrets: All Japanese Secret Weapons, 2008
Imperial Japanese Army Land Weapon Guide, 1997
Japanese Tanks – Tomio Hara, 1978
Japanese Ground Cannons: Heavy Field Cannons – Sayama Jiro, 2012
About the Type 2604/5 on Archive Awareness
“On Japanese heavy tank cooling systems” – Soviet Engineer Grigoriev
The Japanese Superheavy Tanks on For the Record
The O-I on the Warthunder Forums

Surviving Mi-To Blueprint
Surviving Mi-To blueprint – Source

Surviving Mi-To blueprint
Surviving Mi-To blueprint – Source

Mi-To suspension
Mi-To suspension – Source

Mi-To driver's periscope
Mi-To driver’s periscope – Source

Surviving Mi-To track
Surviving Mi-To track – Source

O-I 100 ton
O-I 100 ton – Source

O-I 100 ton artist's interpretation
O-I 100 ton artist’s interpretation – Source

Tank Encyclopedia’s illustration of the O-I 120 ton. Please note the rear turret is incorrectly missing on this model.

Grigoriev's drawings of the Type 2605
Grigoriev’s drawings of the Type 2605 – Source

This picture is sometimes offered as another view of the Type 2604/5 turret-bunker
This picture is sometimes offered as another view of the Type 2604/5 turret-bunker. However, one can tell by looking at it even briefly that it is not the same turret. This turret was located at Kato in Manchuria – Source

Type 91 & Type 95 Heavy
Type 5 Ke-Ho
Share →

17 Responses to Mi-To, O-I, Type 2604 and 2605 Japanese Super Heavy Tanks

  1. 5starWill says:

    Very interesting!
    How did the engineers manage to get the Mi-To out of the mud when it got stuck? Is would seem almost impossible considering the weight of the machine and the lack of other machines to move such large vehicles, unless cranes and tractors were built just to move the new heavies.
    Great article as always.

  2. Seon Eun Ae says:

    I do not appreciate this author stealing my writings and images without giving me credit.

    Also, these “sources” you mention do not mention the O-I and heavies with the information you stole from me. I would ask you to either delete this, or give me the credit without twisting my words. I would otherwise file a plagiarism notice.

    • Stan Lucian says:

      Hello Seon,
      There was indeed a mix-up which led to a couple of sources not having been added. This has been corrected. Also, it was a mistake on our part not to point to the sources of the photos. For that we are sorry. This has also been rectified. Thank you for bringing this to our attention.

      However, as for stealing your work, a review of the two texts has revealed that, while certain paragraphs are based on your work (not copied, based on), our text draws from multiple sources and is also significantly different from your work.

      As for filling a plagiarism notice, if you want to do so, you are free to.
      All the best,
      Stan Lucian, TE Admin

  3. Priest5000 says:

    This is a plagiarized article. You seem to have stolen it from the SENSHA Blog Website (I help the owner of the blog with editing articles), twisted it a bit, and posted it to your own website. You also did not give credit to the SENSHA Blog for any of the information you have copied.

    On the behalf of the writer of the original article, who got her sources straight from the Japanese National Archives and whom did not give you permission to edit her article, I am asking you to please take this down.

    • Stan Lucian says:

      Hello Priest5000,
      As I have responded to Seon above, there was a mix-up with the sources. This is not plagiarism as, while several passages are based on (not copied from, based on) the work of Seon, the article draws up on many sources and is significantly different to anything written on your blog.

      Also, we do not need permission to use publicly available information from the internet. We do have a moral obligation to cite our sources, on which account we have admitted our mistake and corrected it. Any other claims at taking down our ownership are unfounded.
      All the best,
      Stan Lucian, TE Admin

  4. John Tumicki says:

    I had no idea these actually existed. These incredible machines were actually built. On behalf of a tank enthusiast, I believe the O-I exp. in World of Tanks is supposed to be the Mi-To during tests, while the O-I in game is the Mi-To with the additional armor. However, according to your sources, only the 10 cm was mounted. This means if this tank was to become operational, we would have had to face this beast, which I doubt could be defeated until the IS series came about. A scary thought…

    • Seon Eun Ae says:

      Because the author of this article does not know what he is talking about.

      The Tank when under secret planning was labeled as Mi-To. This is an abbreviation of Mitsubishi Tokyo. The tank was designed 150 tons in weight, and having a 15cm Type96 field howitzer. When the tank was ready to be presented during trials, the tank was officially labeled O-I. It weighed 96 tons in prototype form, due to lacking of additional armour palting. That is all there is to it. There is not a 100t or 120 design, not any with a 10cm gun.

      • Hunter12396 says:

        All about the Mi-To you claim is what is written here, except for it being called O-I, I am not convinced of that. As for the O-I (I’m talking about the 120 ton version, I have doubts about the 100 ton variant) not being real, you can take it up with Tomio Hara, great Japanese tank designer and inventor of the suspension type we see on the Type 95 Ha-Go, as he is the primary source on the O-I project.

        But anyway, what are your sources on this subject, I’d really like to see them.

        • Seon Eun Ae says:

          Yes, Hara claims the tank weighed and had a different armament. However according to the O-I report books that were released in 2015, it was shown he only saw the tank for two days. He had no participation in its design and the project as a whole. He is not a credible source in this regard.

          As for my sources, I only rely on the 6 book report series of the tank, its blueprint and other drawings that came with it. All of which I have legal copies of. Whoever wrote this article both uses sources that are outdated and relies on my articles, which I posted on the WT forums as well.

          • Hunter12396 says:

            Even if Hara was not involved in the project in its entirety, how did he come up with very definite details for a tank that are so different from the Mi-To. Hara’s O-I, which you think is the same as the Mi-To, and which I think is a different tank, is so vastly different from the Mi-To; it has different weight, armor, armament, mini-turret configuration, projected speed, and even comes from a different time period! (he claims it’s inception was in 1944 if I recall correctly)

            If the reports you talk about are so groundbreaking, do you care to share them?

        • Seon Eun Ae says:

          Since I somehow cant reply to your latest comment;

          There is only one superheavy tank. This is the O-I tank. The tank was always proposed to have been 150t tons. The prototype without the primary turret weighed 96 tons (It was called 100t by Hara, again, bad memory of his). The turret placed on the chassis and other missing components forced the weight to near 120 tons. You claim you rely on Hara’s words on the tank, which I have seen as well. Here is his words specifically: http://i.imgur.com/HxPn577.png

          According to his book, he has gotten many statistics and details incorrect regarding vehicles. Simply because he eventually became the father of Japanese tanks does not mean he had a hand in everything. He details many times he did not have knowledge to the specific detail on many vehicle projects, In this case the O-I is one of these. He only was presented with the tank during its prototype trials. He was able to examine the tank from afar briefly for two days. Now, he wrote his book decades later and relied on his memory. Because of this yes, many specific stats are incorrect. it is only natural.

          You asked me to share my sources, which come directly from the O-I project report books that FineMoldes purchased in 2015, that were kept held in a shrine. I will not give you large res scans of what I own because I have no interest in you taking my sources and not giving credit like youve done already. But I am going to provide enough to prove to you, that in fact the O-I and “Mi-To” are the exact same project and vehicle.


          When I decide to release them in full resolution, I will have a followup article regarding it. It is natural you were not aware of this, the only primary sources are only for private eyes.

  5. John Tumicki says:

    Oh, yeah. The O-I 100 ton in game is also the O-Ni, That is why it’s weight is 102 tons…

  6. Arizona says:

    Seon, you come off as both a pompous jerk and childish. Considering the fact the author both cited you and was nothing but polite I’d walk your childish comments back. That’s my opinion and your attitude is reason enough for me to boycott your blog as well as look down upon you. Knowledge shouldn’t be hidden unless you yourself authored the report you cite, which you did not.

    Great article TE.

    • Olivier says:

      I’m completely agreeing with you. The person who made these comments apparently can’t manage to see that you need to find sources somewhere, and is apparently very prompt to attack anyone having sources looking similar, but strangely refuses to show where he/she have found her/his own informations. All this, while the moderator has stayed highly polite & professional.

  7. Daboss says:

    Was the type 5 heavy planned to be made or was it the o-I or o-ni

Leave a Reply

Your email address will not be published. Required fields are marked *