Arknights: Endfield gacha — the official pity numbers and how to save

By DayOneFix Editorial · August 24, 2026

Liino’s banner closes at the version handover on 2 September — corrected from 30 August on 24 August 2026, and three days later than we previously said. Yvonne’s rerun opens on 24 September. Between those two dates the only question worth answering is how many attempts you can be holding when the one you want arrives — which is exactly how both player communities frame it. The biggest English video on Endfield pull economy is called “Get 120 Pulls in 1 Day F2P” (Rhinne, about 170,000 views). The biggest Chinese one is 金乌’s 「2天肝出260抽！教你怎麼0氪肝出120抽限定」— two days of grinding, 260 attempts, here is how a free player reaches 120 on a limited (about 1.34 million plays).

Both settle on the same unit, and it is not a percentage. It is 120. Gryphline states it in its own [Official Release Preview] DEV Comm: the Chartered Headhunting guarantee “grants the Rate-UP 6★ operator within the first 120 attempts.”

Everything else on this page is machinery for getting to 120 on the banners you actually want — plus two things nobody can tell you honestly, which are flagged where they come up.

The one trap that costs people limiteds

Endfield does not work like the gacha you came from, and this is the difference that hurts.

Losing the 50/50 does not put you on a guarantee. On a Chartered banner you can pull a 6★, have it be the wrong one, pull another 6★, and have that be wrong too. As Anddad puts it in the most careful English systems breakdown of the banner (about 11,500 views): “unlike other gacha games, if you lose your 6★ 50/50, your next 50/50 is not guaranteed to be the limited character — meaning you can theoretically lose 10 50/50s in a row.” Chinese guides say the same thing in five characters: 「小保底繼承，大保底不繼承」— the small guarantee carries, the big one does not.

The consequence is the whole reason 120 is the planning number rather than 80. Anddad’s conclusion is the mainstream position in both languages: “save at least 120 pulls for a specific limited 6★ — it’s the most simple and straightforward solution, but ensuring you have enough pulls for the 120 guarantee is the best way to limit bad RNG.”

And a corollary people get wrong in the other direction: there is no point deliberately building pity on a banner you do not want. In Anddad’s words, “don’t intentionally build pity, as there aren’t even any five-star characters on rate up anyway.”

Pity, from official notices only

Chartered Headhunting. One published guarantee: the rate-up 6★ within the first 120 attempts. The free 10× from Urgent Recruitment explicitly does not count toward it — Gryphline spells this out, noting that pulling the rate-up inside that free ten “will not affect the guarantee.”

RE-Factor Headhunting, from the [Dreamscape of Wind and Snow] DEV Comm:

• A 5★ or better within 10 attempts.
• A 6★ within 80 attempts, shared across all RE-Factor banners: if you have not hit a 6★, the count carries to the next RE-Factor banner and resets only when you get one.
• The rate-up 6★ within the first 120 attempts, once per banner name-family — so [Resplendent Spectrum] #1 and #2 share it, and the count carries forward. This is the fix for the trap above, and CN spotted it within hours of the preview: 「大保底不繼承終於改了」.
• Free 10× pulls at 30 / 60 / 90 attempts, once per name-family, unused ones carry over.

RE-Factor Issue (weapons), same source:

• No 6★ in 3 consecutive issues means the 4th issue is guaranteed to give one. Does not carry to any other banner.
• No rate-up 6★ in the first 7 issues means the 8th is guaranteed. Once per name-family; the count does carry.
• 10 issues gives an Arms Offering Crate; 18 issues gives the rate-up 6★ weapon outright; after that, every 8 issues alternates between the two.

One thing the RE-Factor rules do not include: a Headhunting Dossier at 60. The official list runs free 10× at 30/60/90, the 120 rate-up guarantee, and a Token every 240 — no Dossier, and no banner-specific expiring permits either. If you are used to farming Dossiers off half-finished Chartered banners, RE-Factor does not pay that.

Two words of warning about vocabulary. English guides do not agree on what to call these numbers: Anddad calls 80 the “hard pity” and 120 “the spark system”, while other creators call 80 the soft pity and 120 the hard pity. Also note that an 80-attempt 6★ guarantee is only published for RE-Factor. For Chartered banners, 120 is the only figure Gryphline puts in writing on the web.

Nobody can tell you how many pulls a version is worth

Two gaps in the official record, and both matter more than they sound.

One: the percentages are not published on the web. Every headhunting notice ends with the same line — “for more headhunting details, check the in-game [Headhunting] screen.” The per-rarity rates are disclosed by the developer inside the client, and we could not locate any Gryphline web publication of them. Chinese press and community coverage consistently reports 0.8% base for 6★, 8% for 5★, 91.2% for 4★, with a soft-pity ramp starting around the 66th pull since the last 6★. Those figures are consistent with the 80-pull guarantee Gryphline did publish, but we label them reported, not verified, and we would not budget on them.

Two: there is no official Oroberyl-per-attempt conversion. We checked every headhunting notice Gryphline has posted since the technical test; not one states a price. The community number is 500 Oroberyls per attempt, 5,000 for a ten-pull — Anddad states it plainly and the major third-party wikis agree — but it appears in no Gryphline publication we could find, and no creator citing it names an official source.

So: every “you can get X pulls this version” claim, including ours, is arithmetic on an unverified constant. That includes the 260-pull and 120-pull headlines that got both communities’ attention. Plan against attempts you can actually see in your account, and against the 120 guarantee, which is real.

What pulling itself pays you

These milestones apply per Chartered Headhunting banner and come straight from the banner notices:

Every pull also pays Arsenal Tickets toward weapon banners: 2,000 for a 6★ operator, 200 for a 5★, 20 for a 4★.

The Dossier is the underrated one. Stopping at 60 on a banner you only half want is not a dead end — it is 10 free attempts on the next one.

The 30-pull counter-strategy, and where it came from

Those milestones are the basis of the most interesting argument in Endfield gacha, which is that saving to 120 is sometimes the wrong play.

The reasoning: 30 attempts buys you a free ten-pull with a guaranteed 5★ or better, 60 buys a Dossier worth ten attempts on the next banner, and every 6★ you pull — rate-up or not — is worth 2,000 Arsenal Tickets. Spread across banners, that can net more total 6★ than funnelling everything into one 120. Endfield Factory’s write-up (about 16,000 views, 1 March) puts numbers on it: “in worst-case scenarios, assuming you lose both 50/50s, you can walk away with three 6-stars across 150 total pulls instead of just two in 120.”

Two things make this worth reporting rather than just noting. First, it is the clearest case of the two communities arriving separately at the same idea: Omega XII published a “30-pull strategy” video on 21 January, and Endfield Factory published the fuller version five weeks later while explicitly crediting the source — “Chinese players have been debating and simulating Endfield’s banner system for weeks, and one strategy came up: the so-called big brain strategy, or 30-120 strategy.”

Second, its own author is honest about the trade: “this strategy is not about maximising banner characters. It’s about maximising total 6-stars.” If you want a specific operator, it is the wrong tool. If you want a deep roster and weapon-banner currency, it beats hoarding. CN has its own compressed version of the same instinct — 「跳三抽一」, skip three banners and pull one.

The rule that decides what to save for

From every Chartered banner notice, verbatim: the rate-up 6★ “will be, after 3 Chartered Headhunting banners, removed from the list of operators that may appear in Chartered Headhunting”, and “will also not be added to Basic Headhunting.”

Three practical consequences:

1. Basic Headhunting will never give you a limited. Do not grind it hoping for one.
2. A limited stays reachable for three banners. Miss Liino’s own banner and she is still in the pool, off-rate, for two more.
3. After that, RE-Factor is the way back — and as of 24 September that is a named, scheduled system rather than a hope. Yvonne, whose banner closed in March, is first.

The banner types

The first [Resplendent Spectrum] RE-Factor Headhunting runs Yvonne rate-up, pool of Yvonne, Ember, Lifeng, Ardelia, Last Rite and Pogranichnik. The paired [Tag Artist Issue] RE-Factor Issue features the 6★ handcannon Artzy Tyrannical.

The currencies behind them:

• Oroberyls — the premium pull currency, for Chartered and RE-Factor Headhunting.
• Chartered HH Permit — usable on any Chartered or RE-Factor banner.
• Banner-specific HH Permits — each banner has its own; up to 5 from a 7-day sign-in event and up to 5 more cheaply through the AIC Quota Exchange. They expire with the banner.
• Basic HH Permit — Basic Headhunting only. Documented sources include both tiers of the Protocol Pass, Regional Stock Redistribution, Etchspace Salvage, and operation-metric milestones.
• Arsenal Tickets — weapon banners and the Arsenal Exchange, where weapons can also be bought outright.
• AIC Quota, Bond Quota, Endpoint Quota — bonus currencies pulling also generates.

Where free Oroberyls actually come from

The [Homecoming] version’s own update notes are the best-documented case, and every figure here is Gryphline’s:

That is 17,600 Oroberyls from one version’s own update notes — our addition of Gryphline’s published figures, before anything repeatable, and before the 100 Oroberyls each Combat Drills operator trial pays. Note that we deliberately stop there rather than converting it into pulls. On top sit the permanent taps:

• Weekly Routine — a permanent screen of weekly missions paying Oroberyls, Arsenal Tickets and materials.
• Exploration crates — the official release raised crate rewards, with upgraded crates containing 150 Oroberyls each.
• [Umbral Monument] and [Delver of the Cryptic] — permanent modes with Oroberyl rewards.
• Protocol Pass — diversified missions and raised rewards at every tier, including Basic HH Permits.

And three outside the client, all officially run:

• SKPORT Daily Sign-in (Game Tools, or the in-game [Protocol Synchronizer]): 300 Oroberyls for your first three daily sign-ins, with limited-window runs paying 260 more over seven days.
• Official socials: up to 180 Oroberyls for following Gryphline’s accounts.
• [Protocol Invitation: The Return]: Oroberyls for inviting returning players, and for returning players who link a code.

A zero-spend plan for one banner

1. Bank the free permits first. A banner’s own 7-day sign-in gives 5; the AIC Quota Exchange sells 5 more at a discount; a Dossier from the previous banner converts into 10. That is 20 attempts before you touch an Oroberyl.
2. Reach 30 for the free 10×, guaranteed 5★ or better, and remember it does not eat into your 120.
3. Reach 60 for the Dossier, which is next banner’s 10 free attempts.
4. Then decide, deliberately. If you want this specific operator, the target is 120, because that is the only number Gryphline guarantees. If you want roster depth and Arsenal Tickets, stopping here and moving on is a defensible play with a real argument behind it.
5. Do not detour into Basic Headhunting hoping for a limited. It is not in there and never will be.

One more thing worth hearing before you reach for a card. Braxophone, mid-way through an Endfield pull session with about 170,000 views, on chasing duplicates past the guarantee: extra copies are close to worthless, and “honestly, don’t really bother whaling on this game — it’s like a 20% damage increase, and I’m not even feeling it.” One player’s read rather than a measurement, but it is the read of somebody who spent the money to find out.

The immediate calendar: the officially named [Dreamscape of Wind and Snow] version opens on 2 September, and Liino’s banner closes with it — the banner has no date of its own, it ends at the handover, so it moved when the version date did. RE-Factor Headhunting begins 24 September with Yvonne. Both are announced rather than inferred, which is rare in Endfield planning; the 30 August we ran before was inferred, and it is superseded.

FAQ

What is the pity in Arknights: Endfield? On Chartered Headhunting, Gryphline guarantees the rate-up 6★ within the first 120 attempts, and that count does not carry to the next banner. On the new RE-Factor Headhunting, the published guarantees are a 5★ or better within 10, a 6★ within 80 (shared across all RE-Factor banners and carried until you get one), and the rate-up 6★ within the first 120, which does carry between banners of the same name.

If I lose the 50/50 in Endfield, is my next 6★ guaranteed? No — and this is the biggest difference from most gacha games. On a Chartered banner, losing to an off-rate 6★ does not put you on a guarantee; you can lose repeatedly. The 120-attempt guarantee is the only thing that ends the run. On RE-Factor banners, the 120 count carries forward between banners of the same name, which softens it.

What are the exact drop rates in Endfield? Gryphline publishes them on the in-game [Headhunting] screen and points every web notice there; we could not find them published on any Gryphline website. Chinese coverage reports 0.8% for 6★, 8% for 5★ and 91.2% for 4★, with a soft-pity ramp from around the 66th pull. Treat those as reported rather than verified, and plan against the official 120.

How many Oroberyls is one pull in Endfield? The community answer is 500 per attempt and 5,000 for a ten-pull, quoted by English creators and the major third-party wikis. No Gryphline notice states a conversion and nobody quoting the figure cites an official source, so we report it as unverified — and it is why “you can get X pulls this version” claims cannot be checked.

Should I save 120 pulls or stop at 30? Depends what you want. If you want a specific limited operator, save 120 — it is the only guarantee. If you want more total 6★ and more Arsenal Tickets, the 30-pull strategy that CN players simulated and English creators picked up can net three 6★ over 150 attempts where 120 nets two. It is a value play, not a way to get the character you actually want.

Do limited operators come back in Endfield? Yes, through RE-Factor Headhunting, which opens 24 September 2026 with Yvonne. Before that system existed, a rate-up 6★ left the Chartered pool after three banners and was never added to Basic Headhunting — still true. RE-Factor is a separate route, not a return to the standing pool.

Is it worth pulling to 60 on a banner I only half want? Yes, on Chartered banners: 60 attempts earns a Headhunting Dossier, which converts into 10 permits for the next Chartered banner once it opens, though those permits expire when that banner ends. RE-Factor banners do not pay a Dossier at all.

How many free pulls can a F2P player get per version? Gryphline does not publish a pull count, and without an official Oroberyl-per-attempt figure nobody can honestly convert one. What is documented is currency: the [Homecoming] version’s own update notes list 17,600 Oroberyls in one-off event payouts, plus repeatable Weekly Routine, Protocol Pass, exploration crate and web sign-in income on top.

Related

• Endfield best characters: who is worth building
• Endfield best teams: the comps and the logic behind them
• The Liino banner and its 2 September deadline
• Endfield 1.5 roadmap
• Endfield codes and how to redeem them
• Endfield 1.5 release time and countdown

Sources

The 120-attempt Chartered Headhunting guarantee comes from Gryphline’s [Official Release Preview] DEV Comm. All RE-Factor Headhunting and RE-Factor Issue rules, the 24 September date, the Yvonne pool and the Artzy Tyrannical weapon banner come from the official [Dreamscape of Wind and Snow] Version DEV Comm, published 22 August 2026 at endfield.gryphline.com, which we read in full; the observation that RE-Factor pays no Headhunting Dossier is our reading of what that rule list omits. Milestone rewards at 30, 60 and 240, the Arsenal Ticket amounts, the currency list and the limited-operator removal rule are quoted from the Chartered Headhunting notices, which repeat them verbatim across all eleven banners. Every Oroberyl figure in the free-pulls table is from the official [Homecoming] Version Update Notes; the 17,600 total is our addition of those published figures. Permanent income sources and the SKPORT Game Tools sign-in come from the official release DEV Comm and Gryphline’s own Game Tools notices. Percentage drop rates are reported from Chinese coverage of the in-client probability disclosure and are explicitly labelled unverified, because no Gryphline web publication of them was located. The 500-Oroberyl-per-attempt figure is community consensus reported by Anddad and the major third-party wikis, with no official source cited by anyone quoting it. Community reads are attributed to their creators and are not our measurement or officially endorsed: Anddad (banner systems, the non-carrying 50/50, the 500 figure), Rhinne and 金乌 (F2P pull-farming framings), Omega XII and Endfield Factory (the 30-pull strategy, the latter crediting Chinese simulation work), Paik (the 120-versus-160 comparison), MooseOnCaffeine (RE-Factor analysis), Braxophone (spending), and Bilibili creators 快乐的逆熵熵 and s我辉煌 on how the guarantees carry. View and play counts were read on 24 August 2026 and move. Publisher and developer per the Steam listing for app 4732690 — Hypergryph and GRYPHLINE — where the game still reads “Coming soon” because Steam is a storefront it has not launched on; official update notices show it live on iOS, Android, PlayStation 5 and PC through the GRYPHLINK launcher, on Asia and Americas/Europe servers, since January 2026.

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.