Arknights: Endfield best teams — how comps are actually built

By DayOneFix Editorial · August 24, 2026

Before you ask which Endfield comp is strongest, ask a better question: which one can you actually pilot? Chinese players have a word for the answer and English guides do not. 「輪椅」— literally wheelchair — is a comp so forgiving it carries you. The most-watched Endfield guide ever made puts it right in the title: 叫我棉被’s launch-week roundup, about 4.4 million plays, is 「這些角色必練！公測隊伍推薦榜…最輪椅配隊有手就行」— “the most wheelchair comp, if you have hands you’re fine.”

English creators reach for the same idea without a fixed word for it. SmugSlav, describing an Ember-led Physical line: “you press counter, you have shields for days, and you can heal. Boom. Like you’re just wheel chairing, man. The game is easy mode.” He also lands on the native version — “just unga bunga, man” — and GodDoggosGacha’s tier list files Laevatain under “comfy and simple that can solve anything.” Same concept, four phrasings, two continents. It is what most people want and almost nobody writes down.

So here are the four comps, sorted by that axis first.

The four comps, and what each one is like to play

These are reported, not measured: they are the compositions high-view Chinese guides built around, with who published them. Names are in the game’s official English forms.

Heat, around Laevatain — the one you can build first and never regret. The line CN settled on is 「萊狼秋羊」: Laevatain as the engine, Wulfgard (5★ Caster) and Akekuri (4★ Vanguard) filling out the element, and Ardelia — a Nature Supporter — in the fourth slot, because Heat does not have a Supporter of its own. 如今生活真的好難’s full build walkthrough is at about 870,000 plays (2 February), and 巴曉白 pitches the same four as 「超低造價萌新之友」— dirt cheap, new-player friendly (about 450,000 plays, 30 January). A WeChat roundup of the top comps summed Laevatain up as 「1托3」, one carrying three: the rest of the slots are whoever you own. 金乌 explains the mechanism in five minutes — she effectively loops her ultimate (about 930,000 plays). English coverage lands in the same place from the other direction: GodDoggosGacha’s tier list calls a Laevatain team “the best team in the game for general use… if you’re looking for something comfy and simple that can solve anything, Laevatain is your girl” (about 18,000 views). Camille changed the ceiling here — the same creator says getting her for Heat “feels like getting Tangtang for a cryo team; the team’s entire performance just shot up.”

Physical, around Pogranichnik — the comp for people who cannot dodge. 巴曉白’s 1.0 guide (about 790,000 plays, 26 January) runs Pogranichnik, Lifeng, Chen Qianyu and the Endministrator, the pure-Physical package the roster predicts. 奈羽落SULUK’s video is titled “a player who can’t use dodge and doesn’t understand boss mechanics opened up the Physical team” (about 480,000 plays) and that is not an insult — it is the selling point. Two warnings from inside the cluster, both worth more than a tier letter: 泰拉構史大王’s 「物理隊最大誤區：你一直在養副C，卻餓死了主C」— the biggest Physical mistake is spreading investment across sub-carries and starving the main one (about 1.0 million plays) — and 月下の幽冥’s argument that the Pogranichnik armour-break rotation everyone copied “may be Endfield 1.0’s single biggest guide misconception” (about 300,000 plays, 10 March). English coverage has since gone further: Paik, working from the Datafield theorycrafting group, says outright that “the main problem is Pogranichnik, because he’s only really efficient against bosses — you need to invest a lot in his breach, and it only works against one enemy” (about 5,000 views, 16 August), and EN tier lists now treat Mi Fu as the Physical core instead. CN barely covered Mi Fu at all. If you play Physical, that disagreement is the most useful thing on this page.

Cryo, around Yvonne and Last Rite — the deepest bench and the hardest hands. 苦力怕到此一遊 called it 「終末地最難操作的隊伍」, the hardest team in the game to pilot (about 1.6 million plays, 13 February), and the same element also has the most-copied lazy rotation in the game. Both are true. YZ-月咒 built the double-carry line — 「冰系夢之隊」, the Cryo dream team — at about 1.57 million plays (28 January); 籽岷 publicly credited him with inventing it. After Tangtang arrived the settled version became 「伊湯賽潔」— Yvonne, Tangtang, Xaihi, Gilberta — which the same creator published as a 「超級懶人循環軸」, a super-lazy rotation, at about 990,000 plays (13 March). Cryo is the only element that can staff all four slots from inside itself, and it shows. The name to notice in that line is Xaihi, a 5★: GodDoggosGacha promoted her to his top tier as “the cornerstone buffer for a lot of cryo teams, if not every single cryo team.” Cryo’s depth is not a 6★ story.

Electric, around Zhuang Fangyi — cheap to start, slow to finish. 巴曉白’s 1.4 guide (about 650,000 plays, 18 July) runs Zhuang Fangyi with Arcane, Perlica and Arclight. There is a 5★ Vanguard in there not by choice: Electric has no 6★ one, and no Defender or Guard at any rarity to fall back on. The honest version of this comp’s case came from a WeChat roundup that listed a 「低成本的電隊」 as one of the game’s four viable lines precisely because you can field it without a 6★ at all — at the cost of damage. Liino filled Electric’s 6★ Supporter slot in 1.4, and the CN framing at the time — 「電訣搶著要」, both the Electric and Arcane camps wanting her — tells you her buff is not Electric-locked. Our Liino build page covers the three things she does not do.

And mono-colour is a habit, not a rule. 叫我棉被 covered Rossi as the enabler of a 「全新混傷隊」— a new mixed-damage comp rather than another single-element line (about 590,000 plays, 1 April).

What English creators rate now, and where they part company with CN

The four comps above are the durable ones — cheap, teachable, still current. The current top of the ladder is a different list, and it has moved twice since launch. GodDoggosGacha’s framing is the useful one: “we’re safely in the second meta now. The first meta would be the heat teams, the cryo teams, with the physical teams. And now the second meta seems to point towards electric and nature teams.”

Two English sources with actual numbers behind them agree on the shape:

• Paik, working from a theorycrafting group he names as Datafield: “there are two teams that are currently stronger than the rest, which are Zhuang Fangyi and Arcane teams… between them, Arcane teams are the strongest.” His lists are Arcane / Liino / Tangtang / Xaihi and Zhuang Fangyi / Liino / Arclight / Perlica, with Liino worth “around 15 to 20%” on the latter. About 5,000 views, 16 August.
• Guobafield, four months earlier: “two of the top teams right now with Zhuang Fangyi, Gilberta, Arclight, Perlica as well as Rossi, Tangtang, Gilberta, Perlica… so which team would I choose to be the best? Zhuang Fangyi just wins.” About 22,000 views, 30 April.

Where that leaves the CN reads: Electric is confirmed from both sides and is the one comp everybody puts at or near the top. Heat and Cryo are agreed on comfort, not ceiling — EN still calls Laevatain the best general-use team while ranking it below Arcane and Zhuang Fangyi for raw output, and Paik’s August read has Yvonne and Last Rite sharing one mid-tier Cryo line. Physical is where the two communities genuinely disagree, per the Pogranichnik note above.

None of that is a reason to distrust the Chinese scene, and the English scene says so itself. Kekvin, in the most-watched EN tier-list comparison of launch week (about 97,000 views): “CN has a head start, or a lot of the knowledge that EN may not have. I value their tier list quite a lot.”

Which operator goes in the fourth slot

The most useful thing anyone has written about Endfield comps is a three-part rhyme, not a tier list. A WeChat guide by 青天小冷面 (2026-02-24) is titled 「順風看潔哥，逆風看小羊，絕境看晝雪」— when you’re ahead, Gilberta; when you’re behind, Ardelia; when you’re cornered, Snowshine. A Bilibili clip from a month later swaps the middle name (「順風潔哥，逆風別禮，絕境晝雪」, 樱小路桃华, 26 March), so treat the middle rung as unsettled — but the two ends are the same in both, and the idea underneath them is the real content.

It is a survivability dial, not a ranking. You are not choosing the best fourth operator; you are choosing how much damage you can afford to trade for not dying, and that changes per stage.

• Ahead — Gilberta. Her battle skill pulls the mobs into a pile and then Last Rite’s battle skill just cuts through them. The article’s use case is open-world exploration and normal-difficulty trash: “kills the whole group at once.”
• Behind — a healer. Ardelia is the Nature Supporter who heals, which is what the title argues for; other players run a second damage dealer here instead. This is the rung people disagree on.
• Cornered — Snowshine. The 5★ Cryo Defender the game handed out at launch. Per the same guide: her battle skill parries an incoming attack, and the shield stays up far longer than a dodge window, so you stop dodging entirely; the parry applies Cryo; the damage reduction is large enough that a full hit barely moves your bar; her combo skill heals for over a thousand; her ultimate freezes. On an event stage that banned tactical items and gave the enemies aggressive attack patterns, the author’s verdict was that Snowshine was “practically invincible.” 「佩麗卡對不起，晝雪她太強了」— sorry Perlica, Snowshine is just too strong.

If you take one thing from the Chinese scene, take this rather than a tier table: swap the fourth slot by how hard the fight is trying to kill you. And note that the operator at the top of the ladder is free.

Why only Cryo can run a mono-element team

The comps above are not arbitrary. Every one of them is a workaround for a hole in the roster, and the holes are not opinion — they are the official operator list, laid out by element and class.

Three things fall straight out:

1. A pure-colour team is only possible in Cryo. Everything else imports at least one slot. When CN guides call a comp 「純色」, mono-element, they are flagging an exception.
2. Nature’s two 6★ Supporters are everybody’s supply depot. Gilberta and Ardelia turn up in Heat comps, Physical comps and mixed comps alike — which is exactly why the fourth-slot rhyme above is built out of them.
3. Physical is the most self-contained core and the most support-starved. Four Guards plus a Vanguard fields damage and frontline with no outside help, and then has nothing of its own to buff it with.

This table is the part of the page that survives new operators. A list of four names is not.

Is the Heat team still the best team?

No, and it has not been for months. 「4大頂級配隊出爐，火隊並非版本唯一答案」— “four top comps are in, and Heat is not the version’s only answer” — is the headline of a widely recirculated WeChat piece from 14 February, and it is the healthiest thing anyone has written about Endfield comps. Launch week produced a Heat monoculture for two reasons that had nothing to do with balance: Laevatain was the launch banner, and she had the biggest guide in the game behind her (叫我棉被, about 2.6 million plays). Within two months Cryo had the deeper bench, Physical had a clean core, and Rossi had opened mixed damage.

The practical version: build the element you already have depth in. Endfield’s progression materials are shared and the roster is small enough that a second full comp is a real cost. The structural table tells you which slot your chosen element will eventually have to borrow.

How rotations are written: combo skills, ultimates and Heat Infliction

Every Chinese guide’s rotation is written in Gryphline’s own vocabulary, so it is worth ten seconds:

• Basic attack — chained sequences. Ember’s is a three-step sequence in the official notes; whoever you leave on the field is doing this.
• Battle skill — the on-demand button. Laevatain’s is Smouldering Fire, with an enhanced version that adds a dashing attack.
• Combo skill — fires when you swap operators, and is the actual glue. Official notes confirm Last Rite’s combo skill freezes, and separately record a bug where it interrupted Tangtang’s ultimate — which is the game admitting those two are meant to be run together.
• Ultimate, fed by Ultimate Energy. Some combo skills grant it and some do not, a distinction the 1.2 notes added to skill descriptions. Laevatain’s ultimate is Twilight.
• Finisher / Final Strike — a separate damage type with its own scaling, referenced for Gilberta, Alesh, Fluorite and Yvonne.

And elements apply statuses — the 1.1 notes name Heat Infliction as a state an enemy is in. A comp exists to keep a status on the target while the damage dealer converts it. That is the whole game.

Where to check comps against the live game

Gryphline runs first-party Game Tools with a Team Picks and Build Guide page where players submit and rate compositions, updated with each new operator. It is at game.skport.com/tools/endfield/rec-team, or in-game through [Protocol Synchronizer] → Game Tools. When a version lands and the comps here age, that is the first place it will show — and unlike a video from January, it moves.

FAQ

What is the best team in Arknights: Endfield? There is no official answer and no single community one. The four with the largest Chinese followings are Heat around Laevatain with a Nature Supporter in the fourth slot, Cryo around Yvonne with Tangtang, Xaihi and Gilberta, Physical around Pogranichnik with Lifeng and Chen Qianyu, and Electric around Zhuang Fangyi with Arcane. Build the one whose members you already own — the roster is small and materials are shared.

What does 「轮椅」 mean in Endfield guides? Literally “wheelchair”: a comp or rotation so forgiving it plays itself. It is a compliment. The related phrases you will see are 「有手就行」(“if you have hands, you’re fine”), 「簡單好抄」(“simple, easy to copy”) and 「高容錯」(“high fault tolerance”). Heat around Laevatain and the Electric line around Zhuang Fangyi are the two most often described that way.

What is the strongest Endfield team in version 1.4? English theorycrafting-backed guides put Arcane teams first and Zhuang Fangyi teams second, both built around Liino. Chinese guides also treat the Zhuang Fangyi Electric line as top-tier. None of it is official, all of it is eight months of a moving roster, and the gap between “strongest” and “the one you own” is usually larger than the gap between the top comps.

Which Endfield team is easiest for a beginner? Heat around Laevatain, on cost as well as difficulty — 巴曉白 pitched Laevatain, Ardelia, Akekuri and Wulfgard as a dirt-cheap new-player comp. Physical is the other easy answer, and it is the one people recommend if you are struggling with dodge timing.

Why does the Heat team use a Nature Supporter? Because Heat does not have one. The official roster lists Heat as Laevatain (Striker), Camille (Vanguard), Ember (Defender), Wulfgard (Caster) and Akekuri (Vanguard) — no Supporter and no Guard at any rarity. Ardelia and Gilberta, both Nature Supporters, are what Heat borrows.

Can you run a mono-element team in Endfield? Only Cryo can staff every class from within its own element. Heat, Electric, Nature and Physical each have at least one class with nobody in it, so their comps import a slot.

How do combo skills work in Endfield teams? Combo skills fire around operator swaps and are the connective tissue of a rotation. Gryphline’s patch notes confirm they can apply status — Last Rite’s freezes — and that some grant Ultimate Energy while others do not, a distinction the 1.2 notes explicitly added to skill descriptions.

Is there an official place to look up Endfield team comps? Yes. Gryphline’s Game Tools include a Team Picks and Build Guide page where players share and rate comps; it updates with each new operator, and is reachable in-game via [Protocol Synchronizer] → Game Tools or on the web through SKPORT.

Related

• Endfield best characters: who is worth building
• Endfield gacha guide: banner types, pity and F2P planning
• Liino build guide, and three things she does not do
• The Liino banner and its 2 September deadline
• Endfield 1.5 roadmap
• Endfield codes and how to redeem them

Sources

Every operator’s class and element, and the element-by-class table built from them, come from the data behind Gryphline’s official Operators page at endfield.gryphline.com, which is also the source for the official English class and element names. Combat vocabulary — basic attack, battle skill, combo skill, ultimate, Ultimate Energy, Finisher, Final Strike, Heat Infliction — and the specific behaviours cited for Last Rite, Tangtang, Ember, Laevatain, Gilberta and Yvonne come from Gryphline’s official version update notes, not from testing. The Team Picks and Build Guide tool is documented in Gryphline’s own Game Tools notices. Snowshine’s status as a launch giveaway is reported by Game8 and by Chinese launch coverage as a global pre-registration milestone reward delivered by in-game mail; we did not find it in a Gryphline news notice. Every comp, nickname and play count is community practice attributed to the creator who published it, and none of it is our measurement or officially endorsed: 叫我棉被 (launch comp roundup, Laevatain, Rossi), 巴曉白 (Physical 1.0, Electric 1.4, budget Heat), 如今生活真的好難 (Heat build path), 金乌 (Laevatain ultimate loop), YZ-月咒 (Cryo double carry, 伊湯賽潔 rotation), 苦力怕到此一遊 (hardest comp), 奈羽落SULUK (Physical forgiveness), 泰拉構史大王 (sub-carry warning), 月下の幽冥 (Pogranichnik armour-break), plus two WeChat guides: 青天小冷面 on 「順風看潔哥，逆風看小羊，絕境看晝雪」(24 February 2026) and 游戏汇总’s four-comps roundup 「火隊並非版本唯一答案」(14 February 2026). The Bilibili variant of the fourth-slot rhyme is 樱小路桃华, 26 March 2026. English-language reads are attributed the same way and are equally unofficial: GodDoggosGacha (per-patch tier lists, the “second meta” framing, Laevatain, Camille, Xaihi), Paik (Datafield-backed comp maths, 16 August 2026), Guobafield (comp comparison, 30 April 2026), SmugSlav (“wheel chairing”, Ember), Kekvin (CN/EN tier-list comparison). View counts are from YouTube on 24 August 2026. Play counts were read on 24 August 2026 and move. Publisher and developer per the Steam listing for app 4732690 (Hypergryph / GRYPHLINE).

Date correction, 24 August 2026. GRYPHLINE’s version overview issued to press on 21 August 2026 dates version 1.5 [Dreamscape of Wind and Snow] to 2 September 2026, reported by Gematsu and carried by Inven Global as a GRYPHLINE press release; it gives a date and no time. Checked 24 August 2026, that date does not yet appear on Hypergryph’s own news hub, whose newest bulletin is still the DEV Comm of 22 August. This page previously said 30 August, which came from a third-party schedule derivation (Game8’s) of when the version would land; the publisher’s own announcement supersedes it, and because the Liino banner is worded as running until “before version update and maintenance”, the banner runs three days longer than we said. Checked 24 August 2026.