Skip to content

Zenless Zone Zero 2.2-2.3 Pull Guide: Who to Pull, Who to Skip

A retrospective pull guide to Zenless Zone Zero 2.2 and 2.3: Lucia, Yidhari, Seed, Trigger and the four reruns rated by roster, plus pity, per-pull costs, worst-case Polychrome totals, and budget planning for each player type.

Gamermir Editor
Published Nov 10, 2026Updated Nov 19, 2026

Banner Status: Ended — 2.2 ran September 4 to October 15, 2025; 2.3 ran October 15 to November 26, 2025 | Updated on: November 19, 2026 | Applies to: Zenless Zone Zero Versions 2.2 and 2.3 (dates UTC+8)

Both versions are closed, so treat this as a retrospective: what each banner was worth, what it cost, and which rosters it fit. That still matters if you are deciding whether a rerun of these agents deserves your Polychrome later.

Quick Verdict

Banner / ItemTypeRecommendationBest For

Lucia

New (2.3 Phase 1)

Recommended

Rupture teams, especially Yixuan owners

Yidhari

New (2.3 Phase 2)

Recommended

Players without a Rupture carry

Trigger

Rerun (2.2 Phase 1)

Situational

Rosters short on Stun

Seed

New (2.2 Phase 1)

Situational

Players with no Electric main DPS

Evelyn

Rerun (2.2 Phase 2)

Situational

Players with no Fire team at all

Orphie & Magus

New (2.2 Phase 2)

Skip

Fire dual-Attack teams only

Vivian

Rerun (2.3 Phase 1)

Situational

Alice owners without a sustain slot

Ju Fufu

Rerun (2.3 Phase 2)

Situational

Rosters missing a Fire Stun

Lucia and Yidhari were the only two banners that could change what a roster can do. Everything else filled a gap you either had or did not have, and Orphie & Magus was the one banner most players could skip outright.

Rating scale, used everywhere below: Must Pull means the best option for a whole archetype and hard to replace; Recommended means a strong pick that pays off for most rosters inside its archetype; Situational means it only makes sense with a specific roster; Skip means the pulls are better spent elsewhere. Nothing here is rated Must Pull outright — every rating assumes a roster, which is why each entry carries its own conditions.

Banners & Schedule

The official 2.2 announcement set the version at 41 days, ending October 15, 2025 at 06:00 (UTC+8); the official 2.3 announcement set 42 days, ending November 26, 2025 at 06:00 (UTC+8), and named Yidhari and Komano Manato as new agents. The phase splits, rerun pairings, and channel names below come from banner trackers, not the official posts — community-reported, and phase switch-over times were never published, so the tables give dates only.

Version 2.2

Dates (UTC+8)ItemBanner TypeRoleKey StrengthPull Priority

Sep 4 – Sep 24

Seed

New agent

Electric Attack

Dual-Attack teams

4

Sep 4 – Sep 24

Trigger

Rerun (1.6)

Electric Stun

Stun utility

3

Sep 24 – Oct 14

Orphie & Magus

New agent

Fire Attack sub-DPS

Dual-Attack synergy

6

Sep 24 – Oct 14

Evelyn

Rerun (1.5)

Fire Attack

Fire main DPS

5

Version 2.3

Dates (UTC+8)ItemBanner TypeRoleKey StrengthPull Priority

Oct 15 – Nov 5

Lucia

New agent

Ether Support

Rupture buffing

1

Oct 15 – Nov 5

Vivian

Rerun (1.7)

Ether Anomaly

Anomaly sub-DPS

—

Nov 5 – Nov 25

Yidhari

New agent

Ice Rupture

Low-HP damage scaling

2

Nov 5 – Nov 25

Ju Fufu

Rerun (2.0)

Fire Stun

Fast ultimate cycling

—

The priority numbers follow one published six-agent ordering; the two 2.3 reruns were not part of it and are left unranked rather than slotted in on guesswork.

Pool composition. Each phase paired its agent with a signature engine channel: Cordis Germina and Spectral Gaze in 2.2 Phase 1, Bellicose Blaze and Heartstring Nocturne in 2.2 Phase 2, Dreamlit Hearth and Flight of Fancy in 2.3 Phase 1, Kraken's Cradle and Roaring Fur-nace in 2.3 Phase 2. Every phase also carried A-Rank rate-ups, but those lists came from a single community source and never appeared in official announcements, so they are not reproduced here. The exception worth knowing is Komano Manato in 2.3 Phase 1, who was also earnable free through 2.3 events and joined the standard Star-Studded Cast from 2.4 — you never needed to pull for him.

Banners running at the same time. The agent channel and the W-Engine channel were live together in each phase and cost the same currency, 160 Polychrome or one Encrypted Master Tape per pull. They closed on the same date, so the engine channel bought no extra time: wanting both the agent and her engine meant paying for both out of one budget inside the same two and a half weeks. Bangboo channels ran on Boopon and the Stable Channel on Master Tapes, so neither competed for Polychrome.

Forward look at the time. 2.4 was reported as Dialyn and Banyue, and 2.5's Ye Shunguang was flagged as a potential top-tier DPS — speculation from a single source, not an official announcement, and the main argument for holding pulls rather than spending them on a marginal banner.

Rates, Pity & Cost

Banner TypeCost per PullBase RateSoft PityHard PityGuaranteeCarries Over?

Agent channel

160 Polychrome or 1 Encrypted Master Tape

0.600% for an S-Rank

Community-reported climb from around pull 74–75, never confirmed officially

90 pulls

First S-Rank is 50/50 for the featured agent; a loss guarantees her on the next S-Rank

Yes, between limited agent banners

W-Engine channel

160 Polychrome or 1 Encrypted Master Tape

Official drop rates are not disclosed

Not disclosed

80 pulls

75/25 in favor of the featured engine; a loss guarantees it next

Not stated in the sources used; engine and agent counters do not mix

Two caveats. The 0.600% and the 50/50 are what rate trackers list for the agent channel — the in-game details screen is the authority and worth opening before you commit. Soft pity is explicitly not official: HoYoverse has never published a starting point or a slope.

What one copy costs. At 160 Polychrome per pull, an agent is 14,400 at hard pity and 28,800 if you lose the 50/50 and repeat it. A signature engine is 12,800 at its 80-pull pity and 25,600 in the 75/25 worst case.

ScenarioPulls NeededCurrencyApprox. USDAssumptions

Early S-Rank, 50/50 won

1–20

160–3,200 Polychrome

TBA

You stop as soon as the S-Rank appears

Typical agent

75–85

12,000–13,600 Polychrome

TBA

Community soft-pity estimate, 50/50 won

Worst-case agent

180

28,800 Polychrome

TBA

Lose the 50/50 at 90, then hit the guarantee at another 90; pity carries between agent banners

Worst-case engine

160

25,600 Polychrome

TBA

Lose the 75/25 at 80, then guarantee at another 80

Dollar figures are marked TBA because no official Polychrome price is part of this guide's sources; convert from your own regional top-up rate before you pull, not after.

Where the pulls come from. One 2.3 count put a full-clear free-to-play player at about 104 pulls, a monthly-pass player at about 120, and pass plus Battle Pass at about 129. A separate, more general estimate for a 42-day version is 11,000–14,000 Polychrome, roughly 70 to 90-plus pulls. The methods differ, the 104 figure is single-source with no stated methodology, and neither is official — plan on 70–104, not the top of the range. Inter-Knot Membership adds 3,000 Polychrome per 30 days, about 18.75 pulls.

Other systems. The Stable Channel gives one free pick from six standard S-Rank agents at 300 cumulative pulls, once per account. There is no fragment exchange and no selector tied to the 2.2 or 2.3 limited banners — the official posts list new agents, engines, and Bangboo only. Limited-banner pity carries over, which is why a failed 50/50 in 2.2 was still worth something in 2.3.

Character Pull Guide

Seed

Pull Recommendation: consider Seed if you have no Electric main DPS and will actually field two Attack agents together. Seed is an Electric Attack agent built to share field time with another Attack agent, a setup the official announcement called out. Before release she was judged likely to out-damage existing Electric options such as Soldier 0 Anby, though that call carried visible uncertainty, and she settled into the upper tier rather than the top of it. Her value is tied to whether you want to build around dual-Attack at all.

  • Pull if: you have no Electric main DPS, or you already run two Attack agents and want the second to be Electric.
  • Skip if: you have a built Electric DPS, or your teams are settled and you are not rebuilding around a dual-carry setup.

Mindscapes: no reliable per-level numbers were available, so M0 is the honest default.

Trigger

Pull Recommendation: pull Trigger only if Stun is the hole in your roster. She is an Electric Stun agent who first ran in 1.6 and returned in 2.2 Phase 1, about six versions later. She ranked third of six on the priority list for these versions — ahead of Seed, behind both Rupture agents — and a tier below them. That is a fair summary: useful, not transformative.

  • Pull if: you are short on Stun agents and one would unlock a team you have already built.
  • Skip if: you have a working Stun option, or you are saving for a Rupture or main-DPS banner.

Mindscapes: no sourced per-level guidance was available; M0 is what the rating assumes.

Orphie & Magus

Pull Recommendation: skip Orphie & Magus unless you already run a Fire dual-Attack team. She is a Fire Attack sub-DPS whose value is concentrated almost entirely in Soldier 0 Anby and Seed dual-Attack lineups; outside those two teams she underperforms for the investment. That is a narrow job, and it is why this is the one banner rated Skip rather than Situational. Soldier 0 Anby also received a 2.3 enhancement, which strengthened the team Orphie & Magus wants — the partner improved after the banner closed, not during it.

  • Pull if: you main Soldier 0 Anby or Seed in a dual-Attack team and want a Fire partner for it.
  • Skip if: you do not run dual-Attack, or you are budget-limited — the same pulls build a team elsewhere instead of completing one here.

Mindscapes: no sourced per-level guidance was available.

Evelyn

Pull Recommendation: pull Evelyn only if your account has no Fire team at all. The veteran Fire Attack main DPS first ran in 1.5 and returned in 2.2 Phase 2. She ranked fifth of six and sits a tier below the Rupture pair, and she is widely described as wanting her signature engine before she feels finished — community opinion rather than a published figure, so treat it as a reason to budget generously, not as a number. By her rerun, most accounts that wanted a Fire DPS had one.

  • Pull if: you have no Fire main DPS and can plan for her engine at the same time.
  • Skip if: you already run a Fire DPS, or your Polychrome is earmarked for the 2.3 Rupture banners.

Mindscapes: no reliable per-level numbers were available; M0 is the baseline.

Lucia

Pull Recommendation: pull Lucia if you run Rupture, and skip her if you do not. She is an Ether Support built specifically for the Rupture archetype — the best buffer available to it by a clear margin, a full tier above the alternatives, with almost no use outside it. One measurement put her buff at +5% HP, +900 Sheer Force, and +30% Crit DMG, roughly 17% ahead of Pan Yinhu and about 35% of team gain with her signature engine in place; the direction is reliable, the exact percentages come from a single source.

Her launch version also brought buffs that made her more forgiving to play: the Dreamscape HP buff went from 30 to 40 seconds, the Sheer Force threshold dropped from 24k to 22k HP, and the Core Passive window doubled from 8 to 12 seconds.

  • Pull if: you own Yixuan or otherwise field a Rupture team and want it competitive rather than functional.
  • Skip if: you have no Rupture roster — she will sit unused — or you still need a damage dealer more than a buffer.

Mindscapes: M1 plus the signature engine was the recommended stopping point in the sourcing used here, on the argument that the pair is where the package becomes worth the pulls. That is one source's call, not a rule; M0 is already a large upgrade over the alternatives.

Vivian

Pull Recommendation: pull Vivian only if you own Alice and still lack a sustain slot. She is an Ether Anomaly sub-DPS who first ran in 1.7 and returned in 2.3 Phase 1. She sits a tier below the headline agents and was rated low priority at the time: the niche is real but narrow, and it assumes a specific partner. One wrinkle pushed her numbers up regardless — the 2.3–2.4 spend incentive for her Iris of the Shore skin counted pulls on any limited banner, so plenty of players reached her total without wanting her.

  • Pull if: you run Alice and need the sub-DPS or sustain slot filled.
  • Skip if: you do not own Alice, or your Anomaly roster is covered — there is no general-purpose case.

Mindscapes: no sourced per-level guidance was available; the rating assumes M0.

Yidhari

Pull Recommendation: pull Yidhari if you want a Rupture carry and do not already have one; otherwise wait and watch. She is an Ice Rupture main DPS whose damage rises as her HP falls — she spends health on charged attacks, claws it back, and can burn Decibels for Ether Veil: Wellspring. The official announcement confirmed the low-HP scaling, and she landed in the upper tier rather than the very top.

This is where the sources genuinely split, and the split is better stated as conditions than as a verdict. With no Yixuan, one argument says pull: she is the fastest route to a working Rupture team. The opposing case is that her element and role overlap with Miyabi and Yixuan, so the slot is crowded, and the better move is to wait and see her settle — pulling only for a specific Ice sub-DPS role, then stopping at M0 plus engine. Both cases assume a different roster.

Her test-server numbers were trimmed before launch: Additional Ability Crit DMG cap from 35% to 30%, M6 Sheer DMG from 30% to 25%, and engine Sheer per stack from 8% to 7%.

  • Pull if: you have no Rupture carry, or you are committing to Rupture as your main archetype.
  • Skip if: you already run Yixuan or Miyabi there, or you would rather see her perform before spending a full pity.

Mindscapes: M0 was the baseline in most sourcing; one assessment suggested M0 plus the signature engine for players who only need an Ice option.

Ju Fufu

Pull Recommendation: pull Ju Fufu if Fire Stun is the missing piece, and skip her if you are only topping up. She debuted in 2.0 and reran in 2.3 Phase 2. The positive case is versatility: her ultimate cycles fast enough that she was called one of the better Stun agents across teams generally, not just Fire ones. The negative case is that a year of roster growth left her looking dated to some players. Both readings agree on the practical point — she is a convenience pick, not a team-defining one.

  • Pull if: you have no Fire Stun and a Fire team whose damage is already in place.
  • Skip if: your Stun slots are covered, or you are pulling her to finish a team that already works.

Mindscapes: no sourced per-level guidance was available for this version.

Weapon / Engine Pull Guide

Every signature engine here belongs to one agent, and the rule is the same for all eight: only pull if you are pulling the character. An engine without its owner has no home, and because the engine channel shares the agent channel's currency and closing date, a bad engine decision eats the pity you needed for the next agent.

Cordis Germina (Seed)

Pull Recommendation: skip unless you are committing to Seed as a long-term carry. It is Seed's signature engine on the Dissonant Sonata channel. No sourced effect breakdown was available for this version, so there is no honest damage figure to quote — only the cost, 160 per pull with an 80-pull engine pity. Free alternatives for the slot were not documented in the sources used.

Spectral Gaze (Trigger)

Pull Recommendation: only pull if Trigger is your main Stun and you are staying on her. Spectral Gaze is her engine on the Vibrant Resonance channel. Its effect numbers were not published in the sources used and no free substitute was documented. Given Trigger's Situational rating, the engine inherits it: an engine for a convenience pick is the easiest place in this guide to waste 12,800 Polychrome.

Bellicose Blaze (Orphie & Magus)

Pull Recommendation: skip unless you already pulled Orphie & Magus and run her in dual-Attack. Bellicose Blaze sits on the Dazzling Choir channel. No effect breakdown was available, and the engine is bound to the one agent rated Skip — which makes it the weakest pull on either version's list.

Heartstring Nocturne (Evelyn)

Pull Recommendation: only pull if Evelyn is already your Fire main DPS and you plan to keep her there. It runs on the Dazzling Melody channel. Evelyn is widely described as engine-hungry, but no sourced percentage for the gap between her signature and a substitute was available, so that reputation is community opinion rather than a measured figure. If you use her, the engine separates a finished team from a workable one.

Dreamlit Hearth (Lucia)

Pull Recommendation: pull Dreamlit Hearth if you already pulled Lucia and intend to keep her on the field. This is the one engine with a measurable case: Lucia's total team gain was put at roughly 35% with the signature in place, and M1 plus the engine was the recommended stopping point for Rupture players. Test-server changes raised its Energy regeneration from 0.24 to 0.5, trimmed its damage bonus from 25% to 20%, and added 12% HP — the final version leans into uptime over raw damage. Those numbers come from a single source's testing.

Flight of Fancy (Vivian)

Pull Recommendation: skip unless Vivian is a core part of an Alice team you already run. Flight of Fancy is on the Vibrant Resonance channel. No sourced effect breakdown or free alternative was available, and Vivian's own rating is narrow, so the engine is a second commitment on an already conditional pick.

Kraken's Cradle (Yidhari)

Pull Recommendation: pull only if you pulled Yidhari and have committed to her as your Rupture carry. It is on the Dazzling Choir channel. Its per-stack Sheer bonus was cut from 8% to 7% during the 2.3 test cycle, so the launched engine is slightly weaker than the version that circulated before release. No free substitute was documented.

Roaring Fur-nace (Ju Fufu)

Pull Recommendation: skip unless Ju Fufu is your dedicated Stun and you are keeping her there long term. Roaring Fur-nace is on the Dazzling Melody channel. No sourced effect numbers or free alternative were available; for a convenience Stun whom some players already considered dated, the engine is hard to justify against a fresh agent banner.

Pull Priority & Budget Planning

Player TypeEstimated Pulls AvailableSuggested Priority OrderNotes

Free-to-play

About 70–104 per version

Lucia, then Yidhari, then stop

The 104 figure is a single community count; budget on 70–90

Monthly pass

About 90–129 per version

Lucia, then Yidhari, then one engine if the agent landed early

The pass adds about 18.75 pulls per 30 days

Heavy spender

Not tracked

Whatever the roster is missing, engine last

Budget by worst case: 180 pulls per agent, 160 per engine

Collector

About 70–104 per version

Reruns first — Trigger, Evelyn, Vivian, Ju Fufu

A missed rerun is the harder gap to close

  • Decide your stopping point before the first pull. Name the agent, the mindscape level, and whether the engine is in scope, then write the number down. The 50/50 is what turns a 14,400-Polychrome plan into a 28,800 one, and that call is far easier before the first ten-pull than after a lost coin flip.
  • Keep one worst case in reserve. Lucia and Yidhari together can cost 360 pulls at worst. Sequence them instead of splitting one budget across both, and let carried-over pity work for you if the first goes badly.
  • Treat the 180-tape skin as a spend, not a bonus. Spending 180 Encrypted Master Tapes on limited banners across 2.3 and 2.4 unlocked Vivian's Iris of the Shore skin. That is roughly one worst-case agent pulled out of 2.4 — worth having only if you were already spending that much, never a reason to start.
  • Set a hard cap and stop at it. Check the rates in-game first, do not top up past a limit you set while calm, and do not let a banner timer decide for you. Reruns happen; none of this is worth borrowing for.

Who Should Pull / Who Should Skip

SituationRecommendationWhy

You own Yixuan or run Rupture

Lucia, then her engine

The archetype's best buffer by a wide margin

No Rupture carry at all

Yidhari, with Lucia if budget allows

Builds the team from zero; the pair is the full package

Already run Yixuan or Miyabi

Skip Yidhari, pull Lucia

The carry slot is filled; the buffer slot is not

No Electric main DPS

Seed, conditionally

Fills the slot, but only if you will use dual-Attack

Rosters short on Stun

Trigger or Ju Fufu, not both

Either covers the hole; a second Stun adds little

No Fire team

Evelyn

The cheapest route to a working Fire DPS

Soldier 0 Anby or Seed dual-Attack main

Orphie & Magus

The one roster where she earns her pulls

Alice owner missing a sustain slot

Vivian

Real niche, but it assumes Alice

New account with nothing built

Skip both versions' reruns

Build one archetype first, then fill gaps

Collection-focused

Reruns over new agents

A missed rerun is the harder gap to close

Conclusion

If you remember one thing: build one archetype properly rather than collecting one good agent from each. Decide from your own roster, not from a ranking, and set your spending cap before the banner opens. Rates are published in-game, a lost 50/50 doubles the worst case, and no agent on this list is worth spending money you did not plan to spend. If none of it fits what you own, waiting cost you nothing.

Update Log

  • November 19, 2026 — Rebuilt as a full retrospective: added the 2.2 and 2.3 phase tables, per-pull and worst-case costs, the W-Engine section, and budget planning. Separated official version windows from community-reported phase dates.
  • October 15, 2025 — 2.3 launched and test-server figures were replaced with launch values: Yidhari's Additional Ability Crit DMG cap 35% 30%, M6 Sheer DMG 30% 25%, engine Sheer per stack 8% 7%; Lucia's Dreamscape HP buff 30s 40s with the Sheer threshold at 24k 22k HP.

More games to download