Azur Promilia Sets Open Beta for November 26: 25 Million Pre-Registrations and the Release-Date Chessboard of Anime Open-World Games#

Note on the lead image: the picture used here is an existing anime open-world themed game image on this site. It serves as a layout illustration only and is not official Azur Promilia material.

Hello everyone, I am Wubian. On October 23, Manjuu Network — the studio best known for Azur Lane — officially announced something the mobile game scene has been waiting on for a long time: the 3D open-world RPG Azur Promilia is set to begin its open beta on November 26, 2026, with the Chinese, Japanese and Korean servers launching on the same day. Two eye-catching numbers came with the date: global pre-registrations have passed 25 million, and the development team has gradually expanded to about 700 people.

As an editor who watches the long-term operation of mobile games year-round, my interest when I first saw this news was not in “another anime open-world game is here,” but in three things: the content-supply pressure behind 25 million pre-registrations, the four-way overseas puzzle assembled by Manjuu, Nexon, Yostar and a Southeast Asian publisher, and just how deliberate the choice of November 26 really is. Put those three together and you basically have an annual case study of the anime open-world race. Let us take them one at a time.

I. First, the Known Facts on the Table: Release Information and Timeline#

As always, let me start by organizing what has been officially confirmed into a table, so nobody has to keep flipping back through news reports.

Now let us walk through the timeline from the end of the third test to the release announcement — the pace is actually quite tight.

A few points worth noting:

• Only about half a month passed between the end of the third test and the release announcement. The test ended on October 7 and the date was announced on October 23, with a round of improvement disclosures squeezed in between — which suggests that adjustment work on the launch version was already scheduled before the test even ended. The date was not picked on a whim; it was back-calculated from the revision schedule.
• The third test was itself a paid test. According to the official line in the interview, the Trail Test was a paid test, meaning validation of the monetization model was already part of the test’s objectives, and pre-launch adjustments to the payment design are being made on the basis of third-test data. As for the conclusions of those adjustments, the official team has only said they are under evaluation and will be announced later.
• The pre-registration figure is stated as “over 25 million,” not an exact value. The official wording is that the milestone has been passed, and the number will most likely keep accumulating before launch. One caveat: this is pre-registration data only. Between pre-registration and the players who actually log in at launch sits the conversion step, and the relevant figures have not been published, so this article makes no estimates.
• The Japanese, Korean and Chinese servers launch on the same day. This point is expanded on in section four; its implications for long-term operation are bigger than they sound.

A few more words on methodology, because the 25 million figure is easily distorted as it spreads:

• The wording is “over,” meaning 25 million is only a lower bound. The official team used “has surpassed” rather than an exact value, pre-registrations will keep accumulating through the final month before launch, and the final number will most likely be higher than this level. When quoting it, keep the word “over” — do not turn it into an exact figure.
• Conversion sits between pre-registration and actual logins. Conversion rates are affected by launch-day server capacity, genre preferences, competition pulling users away and many other factors. The official team has published no methodology, so any equation of “X pre-registrations equals Y day-one players” is unfounded, and this article does none of that.
• Cross-title comparison is not possible for now. There is no authoritative source for the pre-launch pre-registration levels of other recent headweight anime games; the comparison numbers circulating online are all unconfirmed, and this article cites none of them. A cross-title comparison only becomes meaningful once day-one and first-month data start coming out after launch.

II. The Gameplay Foundation: Five Pieces and One Differentiator#

Before we talk about the team and publishing, let me lay out the game’s gameplay foundation, because everything that follows has to rest on it. Based on what has been officially disclosed, the game blends five gameplay modules:

The differentiator most worth noting on top of this foundation is the point the official team keeps stressing: compared with traditional open-world games, Azur Promilia has no mandatory linear progression, and players decide their own itinerary among combat, puzzles, creature catching and homestead management.

Looking at the gameplay list alone, real-time combat plus open-world exploration is already a standard combination in today’s open-world race. What actually gives Azur Promilia a place among comparable titles are the other two pieces: creature collection (Kibo) and homestead management. In other words, the genre-standard features keep it from falling behind, while the differentiators get it remembered. That judgment leads directly to how I read the third-test revision list below — the two heaviest official revisions (the Kibo Duel rework and the gear and creature progression rebuild) both land squarely on the differentiating modules. The direction of the revisions lines up with the positioning.

That sentence may sound like marketing copy, but from a long-term operation perspective it is a structural choice, and it affects several things downstream:

• No mandatory linear progression means handing the decision of “what to do each day” to the player. In traditional open-world games, the daily loop is usually pulled along by a quest chain and the player’s schedule is arranged for them; handing over control of the itinerary means retention no longer rests on schedule lock-in, but on the self-attractiveness of the gameplay modules themselves. Homestead management carries no small expectation on this point — which is why the homestead load reduction deserves to be taken seriously: cutting the load is not cutting content, it is turning the homestead from a “must-do list” into “a place you want to come back to.”
• Running creature catching and character progression in parallel is natural fuel for community topics. Kibo enthusiasm “exceeding expectations” in the third test, on top of the character progression base, leaves far more room for fan creations, guide discussions and collection sharing than a single-track character game does. Whether that ecosystem takes off on its own after launch is the thing I am personally most interested in.

III. A Team of About 700: Manjuu’s Second Curve and the Capacity Yardstick#

Azur Promilia is the biggest genre leap Manjuu Network has made since Azur Lane. Weiwei gave two key pieces of information in his interview with GameGrape: first, that to raise quality and content volume, the team has gradually expanded to about 700 people; second, that Manjuu has now completed the full pipeline from development through self-publishing.

Start with team size. At the level of about 700 people, the meaning in the open-world genre is fairly plain: map output for an open world, production capacity for characters and Kibo, iteration on the homestead system — every one of these is labor-intensive work, and update cadence is strongly correlated with team size. For players, the point of watching this number is to anticipate content supply after launch — whether versions can keep coming after the first month, how long the gaps between major updates are — you can read some of that off the size of this roster. Of course, specific R&D spending, revenue targets and similar figures have not been disclosed, so I will not guess at those.

Then there is the phrase “the full pipeline from development through self-publishing is complete.” Weiwei raised it in the interview alongside team expansion, which shows Manjuu treats publishing capability as supporting infrastructure for this genre leap. A developer that built its name on Azur Lane, moving to developing and publishing its own titles, then carrying that capability into overseas markets — that is exactly the background for the four-party publishing structure unpacked in section four.

Broken down, “the full pipeline is complete” covers at least three capabilities, each with a direct bearing on operations after launch:

• Content supply on the development side. A roster of about 700 backs version iteration for the open-world map, characters and Kibo, and the homestead system. This is the base platform, already discussed above.
• Event and community capability on the operations side. Long-term operation in the open-world genre cannot do without the rhythm design of version events and the upkeep of community topics. Self-publishing means these steps no longer depend on an external agent’s schedule, and the feedback loop between development and operations can be shorter.
• Channel and marketing capability on the publishing side. From pre-registration and pre-download to launch-day traffic, completing the self-publishing pipeline means the domestic line is in their own hands; the overseas division of labor is handled through the alliance structure described in section four.

Together, the three capabilities point to the same judgment: this launch is not a matter of handing the product to a channel and calling it done — Manjuu is building a team around a long-term operation structure. That also matches the posture shown in the third-test revision list: willing to rework what is broken, willing to lighten what is overloaded — both are plays aimed at making content over the long haul.

The revision list published by the official team during the third test is the most direct material for understanding “what changed between the test build and the launch build.” After the Trail Test ended, the disclosed revision directions covered combat, progression and the homestead, as follows:

Two items in this table carry the most weight in my view:

• The Kibo Duel is “reworked,” not “optimized.” The weight of the wording differs: a rework means the third-test version of this system was judged not to work, and the launch version will be a different experience. Kibo is precisely the game’s core signature asset, so how well this item lands directly determines whether the differentiated positioning holds up.
• “Load reduced” is stated explicitly for homestead management. The biggest risk when an open-world game bolts on a homestead system is that it turns into a dumping ground for daily chores. The team cutting the load on its own initiative shows they heard players in the third-test feedback, and also reflects that they have a clear view of where player time should go.

One more unexpected detail from the interview deserves a separate note: Weiwei said the part that “exceeded expectations” in the third test was how much players liked the Kibo and the characters — player enthusiasm for obtaining and raising Kibo was higher than the team expected. From a long-term operation perspective, this matters more than any revision list: if the creature line becomes the focus players naturally gravitate to after launch, it becomes the primary handle on retention and community topics. That said, keep a caveat in mind — gear and creature progression has just been rebuilt wholesale, and whether the rebuilt progression depth can sustain that enthusiasm remains to be verified at launch.

IV. A Four-Party Publishing Effort: How the Overseas Puzzle Fits Together#

The overseas publishing lineup is worth looking at one name at a time: Manjuu, Nexon, Yostar and Southeast Asian publisher PlayPark (SEA) are working together, with the Japanese and Korean servers going live on November 26 alongside the Chinese server. Structure first, interpretation after.

One thing to state up front: the specific regional division of labor and contract details for each of these three overseas partners have not been disclosed, and this article will not speculate. The only confirmed items are the lineup itself and the Japanese and Korean launch dates. Under that premise, there are two layers worth chewing on:

• “The full self-publishing pipeline is complete” does not contradict “finding regional partners.” Self-developed and self-published at home, yet three partners brought in for overseas — this is precisely a snapshot of how Chinese anime games’ overseas playbook is evolving: localization operations, channel relationships, payments and customer support are asset-heavy work that local partners handle more efficiently, while the developer keeps control over version cadence and content quality. Manjuu is building self-publishing capability on one side and assembling an overseas publishing alliance on the other; the two legs serve different markets.
• The operational meaning of launching the Japanese and Korean servers on the same day. A synchronized launch means the Chinese and Japanese/Korean servers start from the same version baseline, and everything downstream — content alignment, event scheduling — has to be done as one global board. For the operations team this is real pressure: a version delay hits all three markets at once. But in the other direction, a same-day launch also concentrates the momentum of the first wave of word of mouth, and anime game users in Japan and Korea are highly concentrated, so community volume in the first two weeks is critical to acquisition that follows.

The most easily overlooked name in the four-party lineup is actually the Southeast Asian publisher. The Southeast Asian market has a large user base and anime game acceptance is still climbing; listing that region separately in the launch publishing alliance shows the structure was designed on a “regional partner” logic rather than simply being bundled and handed to a single global agent.

V. The Scheduling Chessboard: A 50-Day Window Against Ananta#

Here comes the main event. Placed against the recent release calendar of the anime open-world race, November 26 is a remarkably clean move. First, let me put the release dates of several leading titles in the same race side by side:

Now switch to a timeline view and line up the key windows of this year’s race against Azur Promilia’s position:

Games industry outlet GameLook’s read on this date is that Azur Promilia lands 50 days ahead of NetEase’s Ananta global launch on January 15, 2027, while avoiding both the new-title melee caused by the Arknights: Endfield and Neverness to Everness launches in the first half of the year and the anniversary season from September to October when mature titles pile up. I think that judgment covers the main logic of the scheduling strategy, so let me expand on it a little:

• The 50-day gap is a first-mover edge, and also a stress-test window. Opening 50 days ahead of Ananta means Azur Promilia can claim the open-world user pool first, with its initial wave of word of mouth and retention data landing before Ananta enters. But read the other way, those 50 days are also a timed exam: if the first-month data does not hold up, the attention of that same user base will be split when Ananta launches in mid-January. Whether the first-mover scheduling advantage pays off depends on reputation management within those 50 days, not on the date itself.
• Avoiding the first half of the year means avoiding the race’s content arms race. With Arknights: Endfield launching in January and Neverness to Everness in April, the window left for new open-world titles in the first half was already packed. Setting the date in late November puts the pre-launch marketing period right into the lull after summer, with no need to fight those two for exposure, and gives a calmer revision window between the third test and launch.
• Avoiding September to October means dodging the peak daily-activity period of mature titles. Anniversary season is when leading titles hand out rewards and push content hardest; player online time is locked up by older games, and a new title trying to grab attention in that stretch faces difficulty you can imagine. Late November slips past that peak and arrives just before the year-end holiday mood builds — the window left for a new title is narrow, but at least it is clean.

One more layer of race information can be read off this schedule table: within a single year, Arknights: Endfield, Neverness to Everness, Azur Promilia and Ananta enter the field one after another, and the crowding in this race no longer needs an adjective. For Azur Promilia, the scheduling strategy has pushed the odds of a head-on collision with the other three to a minimum; what remains is whether its own differentiation can stand up.

There is one more positional relationship worth spelling out: the first content update cycle after launch will most likely overlap with Ananta’s marketing period. Counting from a November 26 launch and using the standard version cadence of the open-world genre, December through January is exactly the critical window for operating the first version, while the pre-launch marketing push for Ananta will also drive the genre’s overall volume to a high point. Rising volume is good for the race as a whole — the total attention pool grows — but what each title ends up with still depends on its own version content and reputation. That is also why I say the 50-day window is a first-mover edge rather than a moat: the date only decides who speaks first, not how well.

VI. A Watchlist Before Launch#

Finally, here is a checklist of the items I think are worth tracking continuously in the run-up to launch, so you can follow later news against it:

A few more words on the first item. Monetization was the question asked in this interview that did not get an answer — Weiwei only confirmed that the payment design is under evaluation and adjustment and that the third test was a paid test; the specific model has to wait for a later official announcement. Until the official word is out, the various claims circulating about how this game will charge money should not be treated as fact. Please keep that in mind.

The remaining four each have their own points of interest, so let me add a couple of sentences to each:

• How the revision list actually lands (item 2). Reworking the Kibo Duel mechanics, optimizing the capture animation, reducing homestead load, rebuilding gear and creature progression, adjusting rare egg dialogue and the pity system, increasing channels for obtaining high-quality gear — of these six, the Kibo Duel rework and the progression rebuild directly determine how the core loop feels and are the first priority when reviewing the launch version; the homestead load reduction bears on whether the “where should player time go” trade-off discussed above holds up.
• The overseas platform lineup (item 3). The specific platforms on which the game will launch outside mainland China are unconfirmed, and there is no official word on whether a PC client or console platforms are included. For players watching the Japanese and Korean servers, this information directly affects how they get in, so it is worth waiting for official follow-up.
• Synchronized operation of the Japanese and Korean servers (item 4). Opening three markets on the same day makes version alignment and event scheduling no small execution challenge; whether the event schedule in the first week is consistent across all three territories is the most direct window onto how efficiently the publishing alliance runs.
• Pre-registration conversion (item 5). 25 million pre-registrations is the official figure, and no methodology for the conversion rate has been published. The actual scale of players logging in on launch day is the first step in turning the pre-registration number into an operating reality, and it is the moment that tests how much “over 25 million” is really worth.

Closing: Setting a Date Only Writes the Exam Time on the Board#

Bringing this review back to the three points of interest at the top, my conclusion is this: 25 million pre-registrations validate the lower bound of the product’s appeal and content expectations; a team of about 700 and open-world genre version capacity give room to imagine the upper bound; and the four-party publishing plus off-peak scheduling push external competitive variables to the minimum available under current conditions. This move has been played with restraint and clarity from development through publishing to scheduling.

But as the old saying in the games industry goes, setting a release date only writes the exam time on the board — the paper has not been handed out yet. How the reworked Kibo Duel feels, the pace after the homestead load reduction, the final shape of the monetization model — the things that actually decide success or failure will only become clear after November 26. Four leading open-world titles are queued up to enter this race within one year, and none of them has much margin left for trial and error. As someone who watches long-term operation, I will be focusing on creature progression and the social ecosystem in the first month after launch — whether this game’s differentiation card can be played out comes down to that.

See you at launch.