Last Rite build guide — what to do with the operator both languages call the best single-target carry

By DayOneFix Editorial · August 24, 2026

You pulled Last Rite, put her in front of a group of trash mobs, and the game got harder. That is not a build error. Last Rite is the game’s single-target specialist: she erases elites and bosses and has an unusually bad time against packs, and every large guide in both languages says the same thing about her. The fix is not more damage — it is one specific teammate, and knowing which fights to bench her for.

The most-liked comment on any Last Rite video in Chinese puts it better than any guide paragraph. Under 三河尘’s explainer video (about 1.23 million plays), a commenter wrote 「削鐵如泥，但是削泥如鐵」 — she cuts iron like mud, but cuts mud like iron. It has 5,475 likes. 三河尘 replied underneath, 2,052 likes: “fighting bosses feels like fighting trash, fighting trash feels like fighting a boss.” The same line, unprompted, turns up under 梦轩dada’s guide too.

Arknights Endfield operator in combat stance with floating handcannons and teal Etchspace effects — Official Steam screenshot — an operator mid-engagement · © Hypergryph / GRYPHLINE

Two communities, one crown, arrived at separately

This is the rare case where the Chinese and English scenes never talked to each other and landed on the same read anyway — which is worth more than either read alone.

CN’s phrase is 「對單之王」, king of single-target, and it is the title of 梦轩dada’s launch-week guide, about 1.20 million plays on 24 January. 春风满马’s rotation video (about 679,000 plays) argues the mechanism in its own description: every point of her design is spent on single-target burst, and in ideal conditions she has the highest instantaneous single-target burst in the game — but her ceiling depends on teammates supplying the multipliers she does not carry herself.

English creators reached for a different crown for the same operator. Stix calls her an “elite single-target burst specialist” (about 70,000 views). GodDoggosGacha: “queen of burst damage”, “queen of cryo nuke”. XETΛH simply calls her “the Nuke Queen” and titles his optimisation video “P0 Last Rite (Max Damage) Advanced Nuke”. Guobafield’s guide is called “LARGE CRYO DAMAGE!” (about 139,000 views) and says she “deletes single targets” and “struggles in AoE”. DYSPRO’s “P0 Last Rite 686 K Nuke Tutorial” has about 351,000 views — the biggest community Last Rite video in English, and it is a tutorial for one number.

Two scenes, no shared vocabulary, same conclusion. When that happens you can stop shopping for a second opinion.

What is actually official about her kit

Not much, and it is worth knowing exactly how little, because everything else on this page is community testing.

Fact	Where it comes from
6★ Striker, Cryo, entry 19 of 33	Gryphline’s official operator page
Her combo skill freezes enemies	1.1 [Old Deep Water Dies, by Rising Tide It is Denied] update notes, 12 March 2026 — a fix for it “could not freeze enemies under specific circumstances”
Her combo skill can interrupt Tangtang’s ultimate	1.2 [At the Wake of Spring] update notes, 17 April 2026, listed as a bug that was fixed
She is in every Chartered Headhunting pool to date	The banner tables in the 1.1, 1.2, 1.3 and 1.4 update notes

That second-to-last row is quietly the most useful thing Gryphline has ever printed about her: it confirms both that freezing is her mechanism and that she and Tangtang are expected to be on the field together. The plan is freeze the target, then hit the target while it cannot hit back.

Her weapon type is not on the official roster page. CN players describe her as a greatsword user and pair her with Sundered Prince, which the official [Tag Artist Issue] notice lists as a Greatsword — so the category is Gryphline’s, the pairing is the community’s.

The mob problem, and the one operator who fixes it

Here is the practical shape of the complaint, and it is remarkably consistent across thousands of comments.

Under YZ-月咒’s detailed build video (about 837,000 plays), the top explanatory comment — 265 likes — warns: do not touch a Last Rite team early without Gilberta, or you will experience trash mobs, elites and bosses all at the same difficulty, and end up duelling a pack of enemies one at a time. Under 梦轩dada’s guide, a commenter says the same thing from the other side: “I pulled Gilberta, and after that it was easy.”

Gilberta is the Nature Supporter CN calls 「什麼隊都能進的萬用插件」 — the universal plug-in. Her job here is to gather the pack so that Last Rite’s single-target profile stops being a liability. If you are running Last Rite and finding open-world clearing miserable, this is the missing piece far more often than gear is.

The other framing worth stealing, from 悠菜YoNa (about 926,000 plays): she fears several elites on the field at once more than she fears a crowd of trash. Two health bars that both need burst is the worst case, not twenty that need sweeping.

Do not open the world with her

This is the one place where a very large guide’s own author contradicts the impression his video gives, in his own comment section, and it is worth reading twice.

Under 春风满马’s Last Rite rotation guide, the top comment (1,045 likes) is just “this team doesn’t work — I play Cryo, and early progression nearly killed me.” 春风满马 replied, 622 likes: don’t use her for early open-world progression unless you are playing her out of love — the greatsword is too tiring.

The most useful reply in that thread, 360 likes, spells out why: at low investment you have no execute, Last Rite has to stay on the field, the [Trinity] boss will not hold still, so you whiff basic attacks constantly, dodging cuts your own attack chain, and her basic-attack startup and recovery are long enough that two thirds of the fight is spent chasing. Another reply, 359 likes: every guide he watched said the same — even if you pull her, shelf her, progress with the Endministrator and Chen Qianyu, and bring her out later for single-target burst.

The honest dissent is in the same thread, 80 likes: one player used her from day one, took her and her signature weapon to 90, and reports that trash mobs still feel like duels but elites are trivial. He is not disagreeing about the shape — he is saying he was fine with it.

The team is the build

春风满马’s description states the design constraint plainly: she supplies her own high-multiplier skill and her own cold-brittle effect, but everything else has to be filled in by other people, which is why CN calls the standard shape 「3拐1」 — three enablers, one carry.

What the big guides put in those three slots:

Yvonne + Last Rite as a dual-carry Cryo line. YZ-月咒’s build of it (about 1.57 million plays) is titled “development level 100%, the Cryo dream team”. This is the comp most CN players end up in.
Tangtang. 叫我棉被’s Tangtang guide (about 732,000 plays) is explicitly framed as a Yvonne/Last Rite Cryo breakdown, and the official patch note about her ultimate being interrupted confirms the pairing is expected.
Gilberta, for the reason above.
Xaihi, the 5★ GodDoggosGacha calls “the cornerstone buffer for a lot of cryo teams, if not every single cryo team.” A 5★ you can actually finish beats a 6★ you cannot.

And she has not fallen out of the meta. 叫我棉被’s July guide for Arcane (about 448,000 plays) lists a Last Rite/Arcane Cryo comp as one of three recommended lines, seven months after launch.

Three Arknights Endfield operators engaging a large boss enemy in a snowbound arena — Official Steam screenshot — a squad against a large single target, the fight Last Rite is built for · © Hypergryph / GRYPHLINE

Why she is the carry worth actually maxing

The best argument for Last Rite is structural rather than numerical, and a commenter under that Arcane guide made it better than we could (250 likes):

Whoever invested in Last Rite at launch — how far-sighted were you? Back then Cryo ramped too slowly and felt awful. Then Tangtang got a big buff, then Arcane got another. I don’t play any other carry now. And over time I can get her to max Potential.

That last clause is the whole point, and it is official: Last Rite has never left the Chartered Headhunting pool. Every banner you pull for someone else is another chance at a copy of her. Limited operators leave after three banners; she does not. She is the rare carry where “eventually maxed” is a plan rather than a wish.

Gryphline’s [Dreamscape of Wind and Snow] DEV Comm dated the new RE-Factor Headhunting to 24 September 2026, rate-up Yvonne — the other half of the Cryo dual-carry line. Last Rite is in that pool too, as a non-rate-up 6★.

The published rules, in Gryphline’s own words:

5★ or better guaranteed within 10 attempts
6★ guaranteed within 80, and that counter is shared across every RE-Factor banner, carrying over until you get one
The rate-up 6★ guaranteed within the first 120 attempts, once per banner name
A free 10-pull at 30 / 60 / 90 attempts — and those free pulls do not count toward any guarantee or reward total
On ordinary Chartered banners, the rate-up 6★ is guaranteed within 120

Two things nobody can tell you, and you should distrust anyone who does: Gryphline has never published the percentage rates for any Endfield banner anywhere on the web, and there is no official Oroberyl-per-pull conversion. Any pull-cost table you find has quietly invented one of those two numbers.

CN’s read on the rerun banner, from the comment section of 让你爱上学习’s rules explainer (about 77,000 plays), is bracing and worth hearing before you save (144 likes): the rerun exists for exactly two people — someone who does not own her and wants her, and someone topping up Potential. If you are neither, chasing the free 10-pull is worse value than putting those pulls into the next three ordinary banners.

Where the reads genuinely diverge

Fun is not strength, and English coverage is louder about it. Rhinne built Last Rite to max specifically to test his own “most boring character in the game” take, and came back with “she’s by far the most clunky, garbage, boring-ass thing in the game” — while conceding mid-video that “when you actually do damage, it is so much easier to play Last Rite.” If you bounce off her, you are not misplaying her.

English tier lists do not agree with each other about where she sits. Game8’s list, updated 9 August 2026, has her in S and describes her as a single-target carry with strict Cryo Infliction requirements that limit flexibility. Pocket Tactics, updated 1 August, also puts her in S but calls her “the second-best DPS behind Laevatain” — a claim CN largely does not make, because CN treats Laevatain as the comp that plays itself and Last Rite as the specialist. Those are different questions being answered by the same letter.

Guides go stale here faster than they do in most games. The clearest documented example is not even about her: 月下の幽冥 argued that the armour-break rotation everyone copied for Pogranichnik “may be Endfield 1.0’s single biggest guide misconception” (about 300,000 plays, 10 March), and English creators have since moved the Physical crown elsewhere entirely. Anything you copy from launch week deserves a second look.

FAQ

Is Last Rite good in Arknights: Endfield? Yes, at one job. Both the Chinese and English scenes independently rank her as the game’s best or near-best single-target burst carry — CN’s 「對單之王」, EN’s “Nuke Queen” — and both also say she struggles against groups. Gryphline publishes no rankings, so all of that is community reading.

Why does Last Rite feel weak against normal enemies? Because that is her profile, not your build. CN’s most-liked line on her is “she cuts iron like mud, but cuts mud like iron.” Her damage is concentrated into burst on one target, her basic-attack recovery is long, and mobile enemies make her whiff. The common fix is bringing Gilberta to gather the pack.

Who should I put in Last Rite’s team? The most-copied CN comp is Yvonne and Last Rite as a dual Cryo carry line, with Tangtang and Gilberta as the usual support pairing and the 5★ Xaihi as a cheap buffer. The shape CN describes is 「3拐1」 — three enablers around one carry, because she supplies only some of her own multipliers.

Should I use Last Rite to progress through the story? Most Chinese guides say no, including the author of one of the biggest ones, who replied in his own comments that she is too tiring for early open-world unless you are playing her out of love. The standard advice is to progress with the Endministrator and Chen Qianyu and bring her out for bosses.

Is Last Rite permanent or limited? Permanent. She has appeared in every Chartered Headhunting pool published so far, per the official version update notes, and she is also in the RE-Factor Headhunting #1 pool opening 24 September 2026. Limited operators leave the pool three banners after their rate-up; she does not.

How many pulls does Last Rite cost? Nobody can tell you honestly. Gryphline publishes the guarantees — a rate-up 6★ within 120 on Chartered banners, a 6★ within 80 on RE-Factor banners with a shared counter — but has never published the percentage rates anywhere, and there is no official Oroberyl-per-pull conversion. Any specific cost figure you see is someone’s estimate wearing a suit.

Sources

Last Rite’s rarity, class, element and roster position (6★, Striker, Cryo, entry 19 of 33) come from Gryphline’s official operator page, read 24 August 2026; the class and element labels are that page’s own, which is why they are Striker and Cryo rather than the Assault and Ice most third-party sites use. The two kit facts attributed to Gryphline — that her combo skill freezes enemies, and that it could interrupt Tangtang’s ultimate — are quoted from the bug-fix sections of the official “Old Deep Water Dies, by Rising Tide It is Denied” Version Update Notes of 12 March 2026 and the “At the Wake of Spring” Version Update Notes of 17 April 2026. Her presence in every Chartered pool is read from the banner tables of those notes plus the “Sketches of Lost Heirlooms” and “Homecoming” notes. RE-Factor Headhunting, its 24 September date, its pool and every pity figure quoted above come from the “Dreamscape of Wind and Snow” Version DEV Comm of 22 August 2026, read in full the same day. Sundered Prince’s Greatsword category is from the [Tag Artist Issue] pool in that same bulletin. Gryphline has published no percentage rates and no Oroberyl-per-pull conversion in any bulletin we have read; both absences are stated above deliberately. Everything else is community work, reported as opinion with the creator named: Bilibili play counts read 24 August 2026 for 梦轩dada (「對單之王」guide, about 1.20m), 三河尘 (the 4792 explainer, about 1.23m, whose comment section supplies the “cuts iron like mud” line at 5,475 likes), 春风满马 (rotation guide, about 679k, and his own replies in its comments), YZ-月咒 (detailed build, about 837k; the Yvonne/Last Rite dual-carry build, about 1.57m; a 16.63-second boss kill recorded during the third beta, about 1.41m), 悠菜YoNa (about 926k), 叫我棉被 (Tangtang guide about 732k, Arcane guide about 448k), and the comment threads under all of them, which is where the Gilberta fix, the early-progression warning and the long-term Potential argument were found. English reads are attributed to Stix, GodDoggosGacha, XETΛH, Guobafield, DYSPRO and Rhinne, with YouTube view counts read the same day. Tier placements are Game8’s list as updated 9 August 2026 and Pocket Tactics’ as updated 1 August 2026; neither is ours and Gryphline publishes no ranking. The Pogranichnik guide-misconception argument is 月下の幽冥’s, about 300k plays, 10 March 2026. Checked 24 August 2026.