# Kortz Center Heist: calculation dossier

Research checked **4 October 2026**. This is a calculation reference, with a working static planner in `site/dist`. No game client or live tunables were accessed. Guide claims are distinguished from direct source observations and mathematical deductions. Current in-game prices take priority.

**Price-step follow-up:** [PRICE-OPTIONS.md](PRICE-OPTIONS.md) and [price-options.json](price-options.json) derive per-item candidates from GTAWeb's current bounds and a public decompilation of the scoping script. The planner now offers these choices for adding and editing loot, with custom amounts retained. This confirms upward rounding to fixed increments, value-dependent variants, and special egg/fertility duplicate handling. The original image-derived category table below remains provisional: a named gemstone does not inherit its whole category's range, and the image's fertility tier assignments are incomplete. Egg/fertility choices are supersets pending verification of complete sequential generation behavior.

## What the calculator must optimize

Treat each scoped exhibit/case as an individual, indivisible item with an actual price, bag-space cost, location, availability and buyer-request flag. Each of 1–4 players has a separate 100% bag. The primary target is separate from this secondary inventory. A plan must assign every selected whole item to a bag without exceeding its capacity.

The objective is cash after enabled bonuses and costs, not just raw secondary value. Compare an unrestricted loot optimum with an optimum that contains **every requested item**. Buyer completion is all-or-nothing and also required for Elite. Confirm that the entire list has been entered; three entered requests cannot stand in for an unknown fourth. [Scope-out guide](https://www.gtaboom.com/kortz-center-heist-complete-scope-out-guide-d88b).

For a fixed crew size, maximizing host net also maximizes the model's total team payout: these differ only by positive scaling and constants. This equivalence does not hold when comparing different crew sizes.

## Complete secondary catalogue

The attached GTAWeb table supplies these Normal prices and increments. Hard estimates are `Normal × 1.10`, including the increment; apply this only to a Normal estimate. **Never multiply an exact Hard scope-out amount again.** A weekly primary-sale boost does not apply to secondary loot. Catalogue estimates use a midpoint snapped to the stated increment, rather than the maximum. [GTAWeb map](https://gtaweb.eu/gtao-map/ls/-/kortz-center); [GTA BOOM's difficulty/reset observations](https://www.gtaboom.com/gta-online-kortz-center-heist-secondary-loot-payouts-bag-capacity-and-the-best-items-to-steal-6ee7).

| Category | Normal min–max | Normal step | Bag | Hard min–max, derived |
|---|---:|---:|---:|---:|
| Premium gallery paintings | 140,000–162,500 | 2,500 | 50% | 154,000–178,750 |
| Truck crate | 105,000–140,000 | 5,000 | 30% | 115,500–154,000 |
| Gemstones / Venus statues | 100,000–127,500 | 2,500 | 30% | 110,000–140,250 |
| Standard gallery paintings | 102,500–122,500 | 2,500 | 50% | 112,750–134,750 |
| Horse statues / necklaces / skulls | 77,500–100,000 | 2,500 | 30% | 85,250–110,000 |
| Premium fertility statues / meteorite | 70,000–96,000 | 2,000 | 20% | 77,000–105,600 |
| Vault paintings | 70,000–92,500 | 2,500 | 50% | 77,000–101,750 |
| Ornamental eggs / ivory fertility statue | 50,000–64,000 | 2,000 | 20% | 55,000–70,400 |
| Art Deco Circlets | 42,000–53,000 | 1,000 | 10% | 46,200–58,300 |
| Other rings / bracelets | 28,000–35,000 | 1,000 | 10% | 30,800–38,500 |
| Deposit boxes | See uncertainty below | Unresolved | 30% reported; scope unclear | Do not auto-estimate |

All **64 named exhibits**, plus crate and boxes, are in `catalog.json` with their category and evidence references:

| Category | Names / variants |
|---|---|
| Premium paintings (2) | The Great Circle Back; Don't Forgo These Blueprints |
| Gemstones (7) | Emerald; Ruby; Yellow Topaz; Tanzanite; Aquamarine; Gray Spinel; Purple Sapphire — each named “Gemstone” |
| Venus (5) | Venus d'Algernon: Gold; Silver; Bronze; Ivory; Marble |
| Gallery paintings (8) | Sod Off; Cooked; Do You See Me; La Duchesse; Explain Yourself; The Chief; Canis Hominem Edit; Orange Crush |
| Horses (5) | Marble Sabino Criollo; Cremello Dutch Warmblood; Gold Turkoman; Silver Dapple Pinto; Perlino Andalusian |
| Necklaces (6) | Coquard Carcanet: Red Spinel; Emerald; Sapphire; Imperial Topaz; Yellow Diamond; Tanzanite |
| Skulls (6) | Memento Non Mori: Emerald; Ruby; Gold; Amethyst; Sapphire; Diamond |
| Premium small exhibits (5) | Fertility Statue: Mahogany; Gold; Silver; Bronze; Meteorite Fragment |
| Vault paintings (4) | With Friends Like These; Het Gouden Hondje; Swingset Study No. LXIX; The Hunter Becomes the Hunted |
| Eggs (7) | Œuf de Coquard royal; enchanté; décoratif; des abîmes; verdoyant; de vouivre; forestier |
| Standard fertility (1) | Fertility Statue (Ivory) |
| Premium jewelry (1) | Art Deco Circlets |
| Standard jewelry (7) | Coquard Rings; Art Deco Rings; Antique Rings; Coquard Bracelets; Pharaonic Bangles; Byzantine Hoops; Antique Bands |

“Swingest,” “Momento,” “Deposite” and some accent differences in the supplied image have been normalized for search. The separate [Blamp26 reference](https://github.com/Blamp26/kortz-center-heist-calculator) reports these 64 names as extracted from `fm_content_kortz_scoping.ysc` and English localization. That claim was not independently verified against the game files. Its old weights and bonus defaults were not adopted.

## Access, spawn and case interpretation

Two players are needed to open Crisp Gallery. Solo plans exclude its items even if their nominal price is attractive. Reinforced tall cases require the Glass Cutter; deposit boxes require Power Drills. Disable the case alarm for a stealth case-looting route. The loading-bay crate uses a crowbar found nearby; Alpha Mail starts close to it. All these are access/route conditions, rather than price multipliers. [Scope-out guide](https://www.gtaboom.com/kortz-center-heist-complete-scope-out-guide-d88b); [secondary-loot guide](https://www.gtaboom.com/gta-online-kortz-center-heist-secondary-loot-payouts-bag-capacity-and-the-best-items-to-steal-6ee7).

GTA BOOM describes floor slots as follows: bottom 1×10%, 1×20%, 1×30%; ground 5×10%, 1×20%, 6×50%; upper outside Crisp 1×10%, 1×20%, 2×30%; Crisp 1×10%, 2×20%, 2×30%, 2×50%. The vault has up to four secondary painting slots with variable occupancy. The crate is conditional on appearing. These are potential slots, **not a guaranteed available inventory**. Floor numbering varies between guides. The planner initially checks Requires two players for Crisp Gallery categories; confirm this requirement per row. [Secondary-loot guide](https://www.gtaboom.com/gta-online-kortz-center-heist-secondary-loot-payouts-bag-capacity-and-the-best-items-to-steal-6ee7).

The [community wiki](https://w.atwiki.jp/gtav/pages/3237.html) reports that a looted location is empty on the next heist and returns on the following heist, and that requested items can have a higher displayed price. Neither a fixed uplift factor nor a guaranteed request-generation rule has been independently verified here. Record the actual spawn and actual Raf price, rather than applying a buyer-price multiplier.

## Buyer and Elite bonuses

| Payment per player, model | Normal | Hard |
|---|---:|---:|
| Complete buyer list | 50,000 | 100,000 |
| Elite, all conditions met | 50,000 | 100,000 |
| Both | 100,000 | 200,000 |

The solo Hard result screenshots support **100,000 for buyer completion plus 100,000 Elite**, rather than the 150,000 combined figure circulating in early guides. Multiplayer per-player treatment is corroborated by the community settlement model. [Tested payouts](https://www.gtaboom.com/gta-online-kortz-center-heist-payouts-4a8e); [community wiki](https://w.atwiki.jp/gtav/pages/3237.html); [HEISTTAKE](https://heisttake.com/en/przewodnik/kortz-center).

English guides specify under 17:00, accuracy above 90%, loss below 15,000, and every requested item. The Japanese wiki instead gives a 75,000 loss threshold and uses inclusive-looking time/accuracy wording. Check the current game result screen for boundaries and team averaging. The site uses an explicit **“other Elite requirements met” assumption**, not a timer/accuracy prediction. Quick-restart eligibility also needs current in-game confirmation. [GTA BOOM overview](https://www.gtaboom.com/kortz-center-heist-8aff); [community wiki](https://w.atwiki.jp/gtav/pages/3237.html).

## Primary painting and weekly sale

Use the exact planning-board amount. An alternative estimate starts with a Normal repeat-sale base:

`primary cash = sell ? normal repeat base × difficulty multiplier × weekly multiplier : 0`

The weekly multiplier is 4 for the first primary **sale** after reset, otherwise 1; Hard uses 1.10. Keeping/displaying a painting produces no primary sale cash. La Dernière Débauche is the mandatory story target and must be sold; its later availability on Hard is unverified. A pre-reset paid setup can receive the new week's primary-sale boost when sold afterward, without refunding its already-paid fee. [Tested payout guide](https://www.gtaboom.com/gta-online-kortz-center-heist-payouts-4a8e).

| Target | Normal repeat base | Evidence |
|---|---:|---|
| La Dernière Débauche | 481,250 | Normal observed; weekly 1,925,000; Hard availability uncertain |
| Consumato | 308,000 | Normal repeat observed |
| I Hear Voices | 308,500 | Normal repeat observed |
| Mi O Melee | 317,000 | Normal repeat and weekly 1,268,000 observed |
| The Downfall of Rome | 305,000 | Base derived from observed Hard repeat 335,500 / weekly 1,342,000 |
| I, Fruit | 312,000 | Base derived from Hard repeat 343,200 / weekly 1,372,800 |
| The Girl with the Pearl Necklace | 309,500 | Base derived from Hard weekly board 1,361,800; sale unverified |
| The Outcome of Endeavour | 365,000 | Community value; event availability unverified |

Other rotating paintings are supported by the custom-price field. This is **not a verified complete English primary-title catalogue**. A supplemental `primary-community.json` captures 27 Japanese-labelled rows, preserving unverified claims and one detected arithmetic inconsistency. Weekly rotations are date-dependent; do not assume every known painting is available this week. The [community wiki](https://w.atwiki.jp/gtav/pages/3237.html) supplies that larger localized catalogue; the [GTAWeb map](https://gtaweb.eu/gtao-map/ls/-/kortz-center) is a current-rotation starting point, but its dynamic values were not extracted here.

## Settlement model and exact comparison

The corroborated community model allocates the primary to the host and secondary equally among all players. Helpers receive a separate 100,000 Normal / 200,000 Hard base. There are no chosen percentage cuts. Only the host pays setup. [HEISTTAKE](https://heisttake.com/en/przewodnik/kortz-center); [community wiki](https://w.atwiki.jp/gtav/pages/3237.html).

Let `n` be player count, `S` selected secondary value, `L` entered secondary loss, `P` primary cash, `B` buyer bonus, `E` assumed Elite, `V` manually entered per-player vehicle payment, `F` actual host setup fee, and `A` helper base:

```
secondary pool = max(0, S - L)
secondary share = secondary pool / n
host gross = P + secondary share + B + E + V
host net = host gross - F
each helper = A + secondary share + B + E + V
team net = host net + (n - 1) × each helper
```

Do not invent an NPC commission. Whole-dollar display rounding is separate from these equations; actual three-player rounding is unverified. The planner treats loss as a known, fixed amount, not a route-risk model or damage forecast. It does not infer a stealth discount: detection and CCTV may affect real sale value, but no verified penalty equation was found. Input the resulting value directly if known. [GTA Base](https://www.gtabase.com/missions/gta-online/heists/the-kortz-center-heist).

For fixed `n`, a buyer-complete plan with raw value `Sr` beats an incomplete plan `Su` when `(Su - Sr) / n < B + E`, ignoring equal constants and assuming neither pool is clamped by losses. **The bonus does not have to exceed the entire team's lost loot**, only the host's share for host optimization. The corresponding team comparison multiplies per-player bonuses by `n`.

A constructed Normal solo example: unrestricted 374,000 secondary versus buyer-complete 264,000 + 100,000 combined bonuses. Unrestricted pays 10,000 more. With the same entered prices and Hard bonuses, completion pays 90,000 more. This is a mathematical example, not an observed spawn guarantee.

## Deposit boxes: retain uncertainty

The supplied image gives a distribution **conditional on a nonempty box**:

| Value | Probability if nonempty |
|---:|---:|
| 12,500 | 5% |
| 10,000 | 5% |
| 8,000 | 35% |
| 7,500 | 35% |
| 6,000 | 10% |
| 5,000 | 10% |

The conditional expected amount is **7,650** (derived). If `q` is the unknown probability of a nonempty box, unconditional expectation is `q × 7,650`. That is not a guaranteed loot value. The image mentions max 50,000 but explicitly questions whether this means one wall or both. Its 30% weight lacks an unambiguous individual-versus-aggregate scope. GTA BOOM instead quotes 5,000–12,000 and describes 30% for the full set. **Do not multiply per-box value by an assumed box count or allocate 30% per box automatically.** The site requires explicit verification before adding boxes. [Supplied table / GTAWeb](https://gtaweb.eu/gtao-map/ls/-/kortz-center); [secondary-loot guide](https://www.gtaboom.com/gta-online-kortz-center-heist-secondary-loot-payouts-bag-capacity-and-the-best-items-to-steal-6ee7).

## Costs, awards, access upgrades and repeat timing

A repeat setup is reported as 100,000. GTA BOOM reports the first setup started after weekly reset is free; a fee previously paid remains spent. Enter the fee actually paid rather than deriving it solely from weekly sale status. Vehicle payments differ even between similar recorded runs, so default them to zero and allow an actual amount. [Payout guide](https://www.gtaboom.com/gta-online-kortz-center-heist-payouts-4a8e).

Hosting requires a Mansion and Art Studio. Additional properties affect the route: Armory/Bunker improve weapons, vehicle workshop and security help escape, Agency/Garment Factory/Terrorbyte can extend vault time, and an Arcade Drone Station supplies drones. These affect feasibility and time, not a documented fixed loot-value multiplier. Initial property purchase costs are capital costs; keep them out of recurring heist profit unless building a separate break-even analysis. [GTA Base](https://www.gtabase.com/missions/gta-online/heists/the-kortz-center-heist).

The community wiki lists one-time awards: first completion 200,000; first Elite 200,000; Hard without dying 200,000; solo 100,000; four-player completion 100,000. They are not recurring bonuses and are excluded. Career Tier 4 secondary progress reportedly excludes buyer-requested items, making career progress a different objective from cash. Timing claims conflict: the wiki reports a 10-minute replay cooldown and a 48-minute Hard setup window measured from completion; GTA Boss says roughly 15 minutes after Raf's contact; HEISTTAKE uses a 48-minute cooldown. These timings are not simulated. [Community wiki](https://w.atwiki.jp/gtav/pages/3237.html); [secondary-loot guide](https://www.gtaboom.com/gta-online-kortz-center-heist-secondary-loot-payouts-bag-capacity-and-the-best-items-to-steal-6ee7); [GTA Boss](https://www.gtaboss.gg/gta-5-online/guides/kortz-center-heist-finale-walkthrough); [HEISTTAKE](https://heisttake.com/en/przewodnik/kortz-center).

## Conflicts and evidence priorities

1. **Exact scope-out amount wins.** Estimates are transparent and change only when explicitly estimated.
2. **Screenshot step endpoints vs rounded guide bands.** Supplied 102,500–122,500 vs BOOM 102,000–122,000; 162,500 vs 162,000; 127,500 vs 127,000. Do not silently mix those ranges.
3. **Item identity matters.** BOOM labels premium 10% jewelry “Coquard Rings”; supplied table calls it Art Deco Circlets. The site uses the supplied individual name.
4. **Weights conflict.** Early [Steam guide](https://steamcommunity.com/sharedfiles/filedetails/?id=3765266788) and calculator group eggs at 25% and some small exhibits at 10%. Supplied table says eggs/fertility/meteorite 20%, Venus 30%. Overrides are supported.
5. **Hard bonus conflict.** Early 150,000 combined models are superseded here by the later solo result showing 200,000 combined.
6. **Apparent price typos.** [Beebom](https://beebom.com/gta-online-kortz-center-heist-targets/) lists vault paintings at 750,000–850,000. Independent table/guide values are an order of magnitude lower; those defaults were rejected.
7. **Crew settlement remains a community model.** GTA Boss's early flat-fee-only description conflicts with the equal-secondary-share model. Source confidence should remain visible.
8. **Unknown formulas remain unknown.** Buyer-price uplift, deposit aggregation, item spawn probabilities, damage allocation, stealth/CCTV penalty, three-player rounding, exact Elite boundary and complete primary catalogue need further in-game measurements.

The [Rockstar launch page](https://www.rockstargames.com/newswire/article/2525o93834o413/the-kortz-center-heist-now-available-in-gta-online) was located, but returned no extractable text. It is listed as an official reference, not independent numerical confirmation. Many community pages repeat each other; several agreeing articles do not automatically constitute several independent tests.

## Implementation and verification

The browser runs an exact dynamic program over canonical separate bag fills. For each collectible it explores skip/pack actions. A second run forces all buyer requests. Equivalent bag permutations are merged; the highest value per fill state is retained. This is whole-item optimization, not a price-density greedy sort or one bag of size `100 × n`. Weights are editable in 5% increments; arbitrary sub-5% weights and partial exhibit looting are outside this model.

19 automated checks include equipment/access failures, missing requests, per-player settlement, two observed solo payout totals, and 100 randomized inventories compared with an independently enumerated exhaustive assignment oracle. The sample UI also checks exact-price preservation across difficulty changes and impossible solo buyer requests. The recommendation is a packing plan, **not a verified fastest route** or an Elite guarantee.

For additional verification, record the target name, actual location, difficulty, buyer status, Raf price, bag fill before/after collecting, equipment used, crew count and final result screen. Compare the same target across Normal/Hard and buyer/nonbuyer conditions before assigning a new multiplier. Measure deposit boxes individually and as a wall/set before enabling automatic estimates.