Zenless Zone Zero tier list: the Chinese community's 3.1 ranking
By DayOneFix Editorial · August 15, 2026
Below is the Chinese community’s version 3.1 Zenless Zone Zero ranking, transcribed from the chart and mapped to the game’s official English Agent names.
| Tier | Agents |
|---|---|
| T0 |  Ye Shunguang Remielle Velina Dialyn |
| T0.5 |  Ukinami Yuzuha Promeia Aria Yixuan Norma Starlight - Billy Astra Yao Cissia Seed Hoshimi Miyabi Lucia |
| T1 |  Alice Thymefield Yidhari Jane Doe Banyue Orphie Magnusson & Magus Zhao Soldier 0 – Anby Evelyn Chevalier Ju Fufu Burnice White Alexandrina Sebastiane |
| T2 |  Vivian Banshee Qingyi Hugo Vlad Trigger Nekomiya Mana Zhu Yuan Ellen Joe Caesar King Pan Yinhu Soldier 11 Von Lycaon |
| T3 |  Lucy Seth Lowell Grace Howard Koleda Belobog Lighter Piper Wheel Soukaku Asaba Harumasa |
| T4 |  Anton Ivanov Ben Bigger Anby Demara |
Every name on this chart is now mapped. Two Agents that earlier drafts of this page left in Chinese — 琉音 and 希希芙 — resolve to Dialyn and Cissia in HoYoverse’s own bilingual Agent index.
A caution the source itself carries. The creator titles the chart “definitely not dd@ — the streamer’s personal reading”, which is worth taking at face value: this is one informed opinion, not a measured ranking. The value is that it is a Chinese informed opinion, and the arguments behind it are ones English lists do not make.
What the tiers do not tell you
The tiers above are the answer to “who is strong”. The arguments below are the answer to “will she be strong for me”, and in ZZZ they diverge more than in most gacha games — a whole archetype can sink an otherwise fine Agent, and some Agents’ power lives in a partner rather than in themselves.
1. A whole archetype can be quietly deprioritised
The bluntest version, aimed at a top-ten placement:
“She’s only tenth because that whole archetype is weak right now.”
And the advice that follows it, aimed at new players:
“The new class shows that archetype has been abandoned. New players should avoid pulling characters tied to it.”
Whether or not you accept the reading, the mechanic behind it is real: when a game introduces a new class or shifts what endgame content rewards, every character built around the old approach moves — without a single change to their kit.
What to do: when a character sits oddly low, check whether the list is rating her or the shelf she is on. A grade that has quietly become an archetype forecast is the least transferable kind.
The counter-argument is also on record, and worth keeping: “That archetype is declining, but a two-Mindscape character with the right partner still clears everything comfortably.” Investment can outrun a declining archetype for a long time.
2. Sometimes the power is in the partner, not the character
The sharpest single line in these threads:
“At Mindscape 0, most of Miyabi’s strength is actually sitting on Nangong Yu.”
If a large share of a character’s performance comes from a specific partner, then her grade is really a grade for a pairing. Anyone reading it as a solo verdict — and pulling her without the partner — gets a different character than the one the list described.
What to do: for any highly-rated ZZZ character, find out which partner the rating assumes before you budget for her.
3. The opportunity-cost argument
This is the most sophisticated objection in ZZZ discussion, and English lists never make it:
“That team isn’t weak in itself. It looks weak because all three members are among the best partners for another team.”
A squad can be perfectly functional and still be the wrong use of the characters in it. The cost is not the team’s performance — it is what those three characters would have contributed elsewhere. A tier list ranks units; it cannot rank the deployments you gave up.
What to do: if you own a deep roster, rate teams by what they free up, not only by what they clear.
4. Value retention, measured by comparison across versions
A neat heuristic from the same discussion:
“One is from 1.6 and one is from 2.0, and their strength is comparable — that tells you which one holds its value.”
Chinese players routinely compare a character against units several versions older to judge power creep. A new release that only matches something a year old is not the upgrade its launch grade implies; an older character still standing level with new ones is the one that held value.
What to do: treat “is this a real upgrade over what I already own?” as a separate question from “is this good?”, because a tier list only answers the second.
5. When the list itself gets dismissed
Worth knowing the tell, because you will see it: a common one-line reply to a ranking is simply “one look and you know it’s a junk list.” Others dispute specific factual claims — who breaks fastest, for example, gets contested directly, with someone asking whether the creator has actually played the character they ranked below.
That is the healthy part of this ecosystem. A grade nobody can challenge is a grade nobody has checked.
What this page is, and is not
The table at the top is a transcription of a named community chart, not our own ranking — we have not played these Agents and would not pretend to rank them. What we add is the mapping to official English names and the arguments above, which are the part that decides whether a tier applies to your account.
Related
- ZZZ 3.2’s new DEF-scaling class
- Claret Flint, the 3.2 Agent
- Operation Bagel: what to grab and when to leave
- Star Rail tier lists: what a grade hides
- Genshin tier lists: the community keeps score
Sources
The tier chart is transcribed from a Chinese version 3.1 ranking video, captured 16 August 2026; the creator labels it explicitly as a personal reading. Quoted arguments are from viewer comment tracks on high-view Chinese ranking videos for versions 3.0–3.1, translated rather than paraphrased where quoted. Chinese names are mapped against HoYoverse’s own published bilingual Agent index rather than transliterated.