Arknights: Endfield best gear — there is no BiS list, and that is the good news

By DayOneFix Editorial · August 24, 2026

You came here for a best-in-slot table. There isn’t one, and the reason is the best thing about Endfield’s gear: every stat on every piece is fixed and visible before you craft it.

No random main stat. No substat mines. Two identical pieces are identical. What you choose is which set to wear three of, and which attribute you want the numbers pointed at — and once you have chosen, the piece comes out of the factory exactly as advertised. The randomness in this game was moved somewhere else entirely: into Gear Artificing, and into Essences.

Two English comments, on two different guides, land on the same relief. Under IWinToLose, at 15 likes: thank god, a gacha with gear that has set stats and a reliable farm — no more mining for months for crit gear. Under Guobafield’s artificing guide, at 79 likes: I love people complaining about the one RNG system that actually gives you what you want at the end, unlike the 95% of gacha equipment systems that are totally random.

They are both describing the same design decision, and it is the one that makes this page short.

Overhead view of an Arknights: Endfield AIC Factory — production facilities on plots, transport belts, fluid pipes and electric pylons linked by green power beams
Official screenshot: this is the gear shop. There is no other gear shop · © Hypergryph / GRYPHLINE

Your gear comes out of your factory

Endfield’s own store description says the AIC Factory handles “the efficient automated production of various gear, explosives, batteries, tactical items, and other materials.” That is not flavour text. Gear is crafted in Gear Assembly from component items and stock bills, both of which are things your production lines make.

The most honest summary of this came from a Chinese commenter at 2,714 likes, complaining about it: the sad thing about Endfield is how little the base and character progression touch each other — you craft gear and that is basically the whole connection, you cannot craft Essences and you cannot craft ordinary currency or EXP.

He is right, and it cuts the other way too. The one thing the factory does for your combat power, it does completely. An idle base is a closed gear shop. Everything on this page assumes yours is running; if it isn’t, start with why your factory stopped overnight, because none of the rest of this is reachable without it.

A brand-new player reacting to that in real time, under Coping Gamer’s gear-sets guide, at 3 likes: “to get gear, first you have to obtain gear components from the factory” — aight bye, deletes the game. Understandable. Also unavoidable.

The four slots, and the arithmetic that follows

An operator equips four pieces of gear: one armor, one pair of gloves, and two kits.

Rarity runs up a colour ladder to Gold, which is the only tier that matters long-term for one hard reason — only Gold-quality gear can be artificed at all, and artificing is where the last chunk of your power lives. Hypergryph does not spell that boundary out in the version notes, but they do restrict the neighbouring system the same way: the [Homecoming] notes describe advanced assembly techniques as something “Gold-quality gear now supports,” and every guide in both languages puts artificing on the same side of the line.

Set bonuses need three pieces of the same set. Four slots, three-piece bonuses: that means one set bonus and one free slot, always. That arithmetic is the entire “best gear” question, and it is why the answer is a rule rather than a table.

Some of Hypergryph’s own set names, from the version notes rather than from a wiki: AIC Fieldwork, Rift Trekker, Hot Work, Eternal Xiranite and LYNX. Two of those arrived in version 1.4 and three of them gained new pieces in it — which is the first hint that any BiS list has a shelf life measured in versions.

So what actually decides it

Since the numbers are fixed and readable, choosing gear is three questions in order:

  1. Which set bonus does this operator’s job want? Three pieces, one bonus, no hedging. A Defender holding a line and a Striker deleting one want different bonuses, and the bonus is the only part you cannot get any other way.
  2. Which attribute does this operator scale on? Gear carries Defense as its primary attribute across the board; the differentiation is in the attribute bonuses — Strength, Agility, Intellect, Will — and those are printed on the piece before you make it. Match them to the operator, not to the highest number on the screen.
  3. Which stat will you artifice, and can you afford to? That is the only irreversible decision on this page, and it is covered below.

That is it. There is no fourth question, because there is no hidden roll to hope for.

Five Arknights: Endfield operators in Endfield Industries uniforms standing in an industrial quarry area
Official screenshot: the set bonus follows the job, not the operator — which is why one bonus can serve several of them · © Hypergryph / GRYPHLINE

Why we are not printing a per-character BiS table

To be exact about it: Hypergryph publishes no best-in-slot guidance of any kind. Not a recommended set, not a stat priority, nothing. Every ranked gear list in existence is a creator’s opinion.

Those creators are doing real work and some of it is excellent — 梦轩dada’s per-operator 養成作業 series runs into the hundreds of thousands of plays per character in Chinese, and Gacha Gamer publishes an English “best gear, weapons and teams” build video for each new operator. Go and read theirs, and check the version number at the top before you spend anything.

We are not redrawing their tables in English for two reasons. The first is that it would be taking their work and leaving the credit. The second is that these lists genuinely expire: version 1.4 alone added two sets and extended three more, and 1.2’s new sets set off a round of “who is this actually for” videos in Chinese — 梦轩dada’s take on that pair alone (BV1LPddBWEo5) drew about 298,000 plays. A BiS table written against a version you are not playing is worse than no table.

The best comment on this whole topic is under Guobafield’s artificing guide, at 437 likes, and it is the correct amount of respect to give any gear ranking:

In short: bro, why are you worrying about a 0.2% DPS increase — your lead is level 71 and you’re broke.

Gear Artificing: the only irreversible part

Artificing raises a specific stat on a Gold piece. It costs a Gear Artificing Catalyst and it eats another piece of gear as material. It can fail.

The rule that decides your success rate is the one thing everybody gets wrong first, and the clearest statement of it in either language is a Chinese comment at 2,995 likes under 爱画画的暮雨’s five-minute artificing explainer (about 1.12 million plays). Paraphrased:

That last point deserves its own line, because a comment at 272 likes under a related video is just somebody laughing at themselves in public: I spent this entire time thinking the same-named piece was the best one to feed. It isn’t. Go and look for a same-slot piece from a different set with a higher value in the stat you are raising, and make that.

The game labels the good ones for you. Gear whose base value beats your target’s shows as a Good Match; everything else is a Standard Match. And 一升米饭 built an entire video around the fact that you do not need anybody’s lookup table for this because it is shown in the selection screen (about 236,000 plays, titled roughly in case you didn’t know, you can just look at Good Match directly, no table needed).

There is a floor under the whole thing, which is worth knowing before you get discouraged: artificing has a pity. Once a stat’s failed-attempt progress hits its limit, the next attempt is a guaranteed success. GameWith puts the thresholds at 6 attempts for the first improvement, 12 for the second and 20 for the third, and states each stat can be improved a maximum of three times — treat those specific figures as GameWith’s, since Hypergryph publishes no artificing rates, a gap an English commenter under Guobafield’s guide complains about directly.

The two settings quietly costing you materials

One: lock the gear your operators are wearing. A comment at 1,100 likes exists purely as a warning shot — lock your equipped gear, or artificing will happily feed the piece your character is currently wearing into something else. Signed, a man whose fully-artificed piece got melted.

But then read the reply at 219 likes before you go lock everything, because it is the funnier half: locked gear cannot be artificed either, so he locked his whole collection and immediately had to unlock it all again. A commenter at 33 likes has the actual fix — put one level of artificing into a piece and it stops being offered as material automatically.

Two: switch your Gold gear to the newest assembly technique. This one is official, actionable today, and quietly enormous. From the [Homecoming] notes:

“Gold-quality gear now supports switching between different assembly techniques… Newly added assembly techniques are backward-compatible with older ones. When using the latest technique, stock bill consumption enjoys a substantial 99% discount. You can preview compatibility details for all techniques in [Gear Assembly - Switch All Techniques], bulk change assembly techniques for all eligible gear, or customize techniques for individual pieces.”

Ninety-nine per cent, applied in bulk, from a menu. If you have not opened [Gear Assembly - Switch All Techniques] since 1.4 landed, that is the single highest-value click on this page.

For scale on what that discount is against: the Chinese community puts a Gold piece at roughly 8,000 Wuling stock bills and 50 Xiranite components, and notes that melting a mistake refunds about 80% of the stock bills — so the gear itself is close to free and the catalysts are the real cost. One frequently-quoted comment at 406 likes says exactly that: the catalyst is worth more than the gear. Those figures are community accounting, not published prices, but the two languages agree on the shape of it.

In what order, then

The sequence both communities converge on, and the reason this page and the Essence page are separate:

StageWhat you are doingWhy now
1Get the factory running and stop it stallingGear components and stock bills come from nowhere else
2Craft your Gold gear and pick your setsFixed stats, no RNG, nothing to regret
3Cap Exploration LevelEssence yield per run scales with it
4Artifice, then farm EssencesBoth are RNG; only these two are

A Chinese commenter states it as a single instruction: finish exploring, finish the Wuling base, make your Gold gear, then start thinking about graduating Essences. On the English side, a commenter under BladeCube’s guide reaches the same conclusion from the other end — do not touch Essences before the Exploration Level cap, because there are non-random things to spend Sanity on first.

Gold gear is one of those non-random things. That is the whole argument for doing it in this order.

What 1.4 and 1.5 change

Already live (1.4). Switchable Gold assembly techniques with the 99% stock bill discount described above. The [Talosite Exchange] in Wuling trades Talosites for [Gear Templates] among other things. And clearing an operator trial event now hands you an [Arsenal Defense Crate] containing “a special set of Fully Artificed Level 60 gear well-suited for the current trial operator” — with all the matching [Gear Templates] unlocked automatically once you reach Wuling. A free, finished, correctly-matched set is not a bad way to learn what a finished set looks like.

Previewed, not shipped (1.5). The [Dreamscape of Wind and Snow] DEV Comm of 22 August 2026 lists two gear features under “Optimization Content Still in Development” — meaning they are an early preview of planned content, not a 1.5 launch feature. Plan accordingly:

That last clause is the tell. Hypergryph has watched people over-assemble fodder for a stat that then succeeded on the second attempt, and is building the refund.

Is it worth hoarding catalysts until then? An English commenter asks that exact question — whether to hoard against a hypothetical level 80 or 90 gear tier. No tier above Gold appears in any official version notes through the 1.5 DEV Comm. You cannot prove a negative about unannounced content, but “wait” currently rests on nothing published, and the cost of waiting is charged to you every day you fight with unfinished gear.

FAQ

What is the best gear for each character in Endfield? There is no published best-in-slot list, from Hypergryph or anyone with authority. Because gear stats are fixed and readable before crafting, the decision collapses to a rule: pick the set bonus that suits the operator’s job (you get exactly one, since bonuses need three of your four pieces), then pick pieces whose attribute bonuses match what the operator scales on. Per-operator build videos exist from creators in both languages and are worth watching — check which version they were made for.

What gear slots are there? Four: one armor, one pair of gloves, and two kits.

Do I get a set bonus from two pieces? No. Set bonuses require three pieces of the same set, and you only have four slots — so one set bonus and one free slot is the maximum configuration.

Are gear stats random? No. Every stat is fixed and visible before you craft the piece; two copies of the same gear are identical. The randomness in Endfield’s progression lives in Gear Artificing and in Weapon Essences, not in the gear itself.

Which gear can be artificed? Only Gold-quality gear. Hypergryph’s own [Homecoming] notes describe the advanced assembly techniques as something “Gold-quality gear now supports,” and every guide in both languages agrees the Gold boundary applies to artificing too.

What is a Good Match in Gear Artificing? A material piece whose starting value in the stat you are raising is higher than your target piece’s, which gives a better success chance than a Standard Match. It has to be the same slot, and it has to carry the same stat — but it does not have to be from the same set, which is the mistake most people make once and remember forever.

Does artificing have pity? Yes. Once a stat’s progress hits its limit the next attempt is guaranteed to succeed. GameWith puts the thresholds at 6, 12 and 20 attempts for the first, second and third improvement, and caps improvements at three per stat; Hypergryph publishes no artificing rates, so treat those specific numbers as that site’s.

Why did artificing eat the gear my character was wearing? Because it was not locked. The counter-warning matters too: locked gear cannot be artificed, so locking everything is not the answer. Putting one level of artificing into a piece stops it being auto-selected as material.

Am I wasting stock bills on gear? Possibly, and there is a menu that fixes it. Since 1.4, Gold gear can switch assembly techniques, and using the newest technique cuts stock bill consumption by 99%. You can bulk-apply this from [Gear Assembly - Switch All Techniques].

Should I hold my catalysts for a higher gear tier? Nothing published supports that. No tier above Gold appears in any official version notes through the 1.5 DEV Comm of 22 August 2026, while the cost of fighting with unfinished gear is charged to you today.

Gear or Essences first? Gear. It is deterministic, it comes from your factory rather than your Sanity, and it has no daily clock on it — an English commenter makes the point that gear can simply be pocketed until you decide, which is exactly what takes the pressure off. Essences are the daily Sanity spend, and their yield scales with your Exploration Level, so the sequence is: base, then Gold gear, then Exploration cap, then Essences.

Sources

Official. The Gold-quality assembly technique switching, its 99% stock bill discount and the [Gear Assembly - Switch All Techniques] screen; the gear sets [AIC Fieldwork], [Rift Trekker], [Hot Work], [Eternal Xiranite] and [LYNX]; the [Talosite Exchange] route to [Gear Templates]; and the [Arsenal Defense Crate] with its “Fully Artificed Level 60 gear” and auto-unlocked templates are all quoted from the Homecoming Version Update Notes and the Homecoming Core Chapter Version DEV Comm. The [Gear Plan] feature and the Auto-Artificer tab with its material-return behaviour, together with their placement under “Optimization Content Still in Development”, are from the Dreamscape of Wind and Snow Version DEV Comm of 22 August 2026 — which is also the document in which no gear tier above Gold appears. “The efficient automated production of various gear, explosives, batteries, tactical items, and other materials” is from the Steam store listing, which credits Hypergryph as developer and GRYPHLINE as publisher. Operator classes (Striker, Defender, Guard, Caster, Supporter, Vanguard) and elements (Heat, Cryo, Electric, Nature, Physical) are read from the localisation bundle behind the official Arknights: Endfield site. Hypergryph publishes no artificing or engraving rates, no stat priorities and no best-in-slot guidance of any kind; everything in this page’s decision rule is arithmetic on published mechanics, and everything else is attributed.

Community, named and counted. Play and view counts measured 24 August 2026 and rounded. The Good Match rule — same slot, same stat, material value not lower, higher value giving a success bonus, and no same-set requirement — is the top comment at 2,995 likes under 爱画画的暮雨’s 「五分钟看懂装备精锻」 (BV1L66wBNESC, ~1.12M plays), cross-read against 靡弦’s mechanics one-pager (BV1kG6jBSEHf, ~997K plays) and, on the English side, WG:B’s full artificing guide (~84K views) and Guobafield’s (~117K views), whose chapter list is itself the source for the armor / gloves / kit split. The in-game Good Match display is 一升米饭’s whole point (BV1ZNAqzsESD, ~236K plays), where the same-name misconception appears as a comment at 272 likes. The locking warning at 1,100 likes, the locked-gear counter at 219 likes and the one-level-of-artificing fix at 33 likes are from 靡弦’s and 一升米饭’s threads. Gold gear costing roughly 8,000 Wuling stock bills and 50 Xiranite components, the ~80% refund on melting and “the catalyst is worth more than the gear” are from a comment at 406 likes in 一升米饭’s thread, with the stock-bill figure independently reported in English coverage; treat both as community accounting. “Your lead is level 71 and you’re broke” is a comment at 437 likes under Guobafield; the fixed-stats relief is comments at 15 likes under IWinToLose (~234K views) and 79 likes under Guobafield; the “pocket the gear, no expiry” argument is a comment at 108 likes under Guobafield; the missing published rates is a comment at 33 likes there; the higher-tier hoarding question is a comment under Paik’s endgame guide (~67K views). The factory-is-the-only-link complaint at 2,714 likes is from 泥萌都是托’s thread (BV1e2zDBhEAe, ~1.73M plays); the “deletes the game” reaction is under Coping Gamer’s gear sets guide. Ordering — base, then Gold gear, then Exploration cap, then Essences — is cross-verified between a Chinese comment under 爱画画的暮雨’s Essence explainer (BV1swzQBnEZo) and a comment at 34 likes under BladeCube (~131K views). Artificing pity thresholds are GameWith’s and labelled as such in the body; the four-slot layout and the three-piece set rule are corroborated by the Endfield Talos Wiki, whose set naming matches Hypergryph’s own where the two overlap.

Boundaries. No per-character best-in-slot table appears on this page, because none is published by Hypergryph and the ones creators publish are version-scoped opinion that belongs to them — we name where to find them instead. GameWith’s artificing thresholds and the community’s gear-cost accounting are labelled as their sources’ claims in the body, since they are the numbers a reader would budget materials around. No blueprint layout, blueprint code, facility placement, coordinate or build order from any creator is reproduced here. All sources checked 24 August 2026.