How to cite
Source: aion2builds.com armory sample, 2026-10-16, n=1,600
Download the data as CSV. Link to https://aion2builds.com/stats/ when you use it.
How we sampled
Official-sample n=1,600, 2026-10-16
- For each class, level-45 characters were collected through the public character search (all gender and race variants, keywords a, first result pages, sorted by level, all five regions) with an equal per-region quota.
- Character info was read for every candidate, and the top 200 per class by combat power were analysed in full (equipment, skills, stigmas, titles, wings, pets).
- Daevanion board structure is the same for every player, so it was read from 2 characters per class.
- Item detail comes from the official item endpoint (base stats at enchant 0 plus the most common enchant level).
- Only aggregated numbers are published. No player names, character IDs or profile images are stored in these files.
- Region mix is set by the sampling quota (equal per region), not by real player distribution.
- Sample bias: search order within level 45 is not random, and the top 200 are chosen by combat power, so results describe strong level-45 characters, not all players.
Sample size
Sampled on 16 October 2026. The top 200 per class by combat power get the full read. The columns below count the characters behind each class.
| Class | Level 45 candidates | Top sample (n) |
|---|---|---|
| Assassin | 400 | 200 |
| Chanter | 400 | 200 |
| Cleric | 400 | 200 |
| Gladiator | 400 | 200 |
| Ranger | 400 | 200 |
| Sorcerer | 400 | 200 |
| Spiritmaster | 400 | 200 |
| Templar | 399 | 200 |
Combat power by class
Combat power (CP) of each class's top 200. The bar runs from the 10th to the 90th percentile, the tick marks the median, and the dot is the highest value.
| Class | p10 | Median | p90 | Highest | Median, all sampled |
|---|---|---|---|---|---|
| Chanter | 64,030 | 72,758 | 99,260 | 136,505 | 61,113 |
| Assassin | 59,722 | 70,032 | 98,286 | 123,897 | 56,852 |
| Sorcerer | 61,129 | 70,493 | 97,794 | 118,927 | 58,840 |
| Templar | 60,498 | 70,588 | 93,529 | 116,053 | 57,721 |
| Cleric | 62,583 | 72,349 | 91,749 | 151,737 | 60,893 |
| Gladiator | 60,586 | 70,469 | 91,533 | 134,737 | 58,403 |
| Ranger | 60,845 | 71,405 | 90,616 | 139,580 | 58,604 |
| Spiritmaster | 53,230 | 67,602 | 90,003 | 135,573 | 51,112 |
Share of the top sample
We pooled every sampled character, then looked at the strongest 100, 500 and 1,000 by combat power. Each bar is the class's share of that group. With equal sampling per class, 12.5% would be an even split of eight.
| Class | Top 100 | Top 500 | Top 1,000 |
|---|---|---|---|
| Chanter | |||
| Cleric | |||
| Ranger | |||
| Templar | |||
| Sorcerer | |||
| Gladiator | |||
| Assassin | |||
| Spiritmaster |
Regions and races
The region mix of the top 200 per class. The region quota is ours, so it is not the real player split. Race is the split inside the top sample.
| Class | NA East | NA West | Europe | Latin America | Asia | Race |
|---|---|---|---|---|---|---|
| Assassin | 16% | 24% | 27.5% | 14.5% | 18% | Elyos 55%, Asmodians 45% |
| Chanter | 7.5% | 29% | 20% | 14% | 29.5% | Asmodians 50.5%, Elyos 49.5% |
| Cleric | 10% | 27% | 17% | 22% | 24% | Asmodians 53.5%, Elyos 46.5% |
| Gladiator | 12.5% | 25.5% | 23% | 15% | 24% | Asmodians 50%, Elyos 50% |
| Ranger | 13% | 26.5% | 23% | 13.5% | 24% | Elyos 51%, Asmodians 49% |
| Sorcerer | 12.5% | 24.5% | 24.5% | 13% | 25.5% | Elyos 55.5%, Asmodians 44.5% |
| Spiritmaster | 10% | 33.5% | 23.5% | 18.5% | 14.5% | Asmodians 58.5%, Elyos 41.5% |
| Templar | 10.5% | 28.5% | 22.5% | 21.5% | 17% | Elyos 55.5%, Asmodians 44.5% |
Daevanion progress
The median share of each Daevanion board that characters in the sample have opened. Boards unlock over time, so expect these to rise with each patch.
| Class | Nezekan | Zikel | Vaizel | Triniel | Azphel |
|---|---|---|---|---|---|
| Assassin | |||||
| Chanter | |||||
| Cleric | |||||
| Gladiator | |||||
| Ranger | |||||
| Sorcerer | |||||
| Spiritmaster | |||||
| Templar |
Caveats
- Combat power follows gear. It is not a skill or balance score.
- The sample is strong level 45 characters. It does not describe casual players.
- Each class gets the same quota, so shares show relative strength, not popularity.
- Candidates come from public search order, which is not random.
- A patch can change every number. Check the sample date.
Every class is sampled with the same number of level-45 characters, so these are NOT population shares. "shareOfHighCpSample" says which classes supply the highest-CP characters among equal-size per-class samples; read it as relative strength, not popularity. Region shares are also set by the equal per-region quota and are not the real player distribution. Candidate names all contain the search letter, and only the first result pages by level were read.
Steam peak players
AION 2 peaked at 397,905 concurrent players on Steam on 5 October 2026, the day of the Global launch. The figure is Steam only. Players on PURPLE, NC's own launcher, are not counted.
Source: Inven Global, 5 October 2026, citing SteamDB. We have not read SteamDB ourselves.
Stats FAQ
Where does the data come from?
Public official AION 2 character pages. We read level 45 characters through the public search and character endpoints, at no more than one request per second, and keep only aggregated numbers.
Does this show how many people play each class?
No. Every class is sampled with the same quota, so shares show relative strength among top characters, not population.
Why use combat power?
It is the one number every character page shows. It follows gear, so read it as progress, not skill or class balance.
Are player names stored?
No. We publish aggregates only. No names, IDs, guilds or server names are published.
Can I reuse the numbers?
Yes. Cite the source line in the box above and link to this page. The CSV has the same figures.
How often does it update?
We refresh the sample on a schedule and show the sample date on every figure.