Consensus
Source matrix
| Class | AION2 Tier List · Oct 12 | Mein-MMO (DE) · Oct 13 · best role tier | Gameblog (FR) · ~Oct 9 | AllThings.how · Aug 30 [KR/TW] | LagoFast (zh-TW) · Sep 23 [KR/TW] | Consensus | Agreement |
|---|---|---|---|---|---|---|---|
| Gladiator | A | A | S | S | S | S | 60% |
| Templar | A | S | B | A | S | A | 40% |
| Assassin | A | A | A | B | A | A | 80% |
| Ranger | A | A | S | S | A | A | 60% |
| Sorcerer | A | A | S | C | B | A | 40% |
| Spiritmaster | S | A | S | D | A | A | 40% |
| Cleric | A | S | B | A | S | A | 40% |
| Chanter | S | A | A | A | S | A | 60% |
Global 1.0Not confirmedOct 15, 2026
How it is calculated
- Formula
- Consensus tier = median of all source tiers on a common S–D scale; agreement = share of sources that put the class exactly in the consensus tier.
- Data source
- 5 public tier lists (linked in the table header)
- Data version
- Global 1.0 · Oct 15, 2026
- Confidence
- Medium — tier lists are opinions; the more sources agree, the more reliable the tier
Editorial verdict
- Gladiator A
- Consensus S (3 of 5), but both Global meter-based lists (AION2 Tier List, Mein-MMO) say A. The S votes come mostly from KR/TW-based lists. We rate A until Global data agrees.
- Templar A
- Consensus A. Every group needs a main tank, and Mein-MMO rates it S as a tank. As a damage source it is weaker, so A fits.
- Assassin A
- The most stable rating: A in 4 of 5 lists. Strong single-target burst, weaker in long fights.
- Ranger A
- Consensus A (3 of 5). Safe, consistent ranged damage; S only on two lists.
- Sorcerer A
- Widest spread (S to C). AION2 Tier List ranks it second on Global dungeon meters, but it needs gear. A is the fair middle.
- Spiritmaster A
- Most disputed class: S on two lists, D on one KR/TW list. Global meter data looks good, but the sample is small. A until it settles.
- Cleric A
- Consensus A. The main healer is essential for hard content, which most lists value more than its own damage.
- Chanter A
- Consensus A (3 of 5). AION2 Tier List rates it S on Global meters; one source is not enough to raise it yet.
Methodology
Meta tier list. We read public tier lists in 4 languages and keep only lists that rate at least 6 of the 8 Global classes. Each list is converted to one S–D scale: letters are kept as-is; 3-tier scales (T0/T1/T2) map to S/A/B; custom scales are mapped as described in our source matrix page. The consensus tier is the median of all ratings (on an even count we take the better of the two middle values). Agreement = how many sources gave exactly the consensus tier. Column labels show the data basis: [KR], [TW] or [KR/TW] marks lists built on the Korean or Taiwanese service, which is ahead of Global (level cap 50, Brawler). Partial lists that only name top picks (Game8, GamesFuze) are cited on the source matrix page but are not part of the consensus. The editorial verdict is separate from the consensus and gives more weight to lists that use Global data. Player voting is separate again. Snapshot: Oct 15, 2026.
Player vote
Player voting is shown separately from the source consensus and the editorial verdict.
