BATTLEFIELD 6 + REDSEC · WEAPON SCIENCE

Engineer the
perfect loadout.

Build around how you play. Explore every compatible attachment, see the tradeoffs, and find the right balance of control and ideal kill timing.

56 WEAPONS INDEXED DATA PATCH3 SEARCH GOALS

LIVE BUILD READOUTSYNCED
MANUAL PREVIEW
PRIMARY

Choose a weapon

PROFILE RATING—
POINTS—

Dashed = baseline · cyan = current preview. Profile rating is the mean of the six capped display axes, not your optimizer score. Changes describe this preview versus the weapon baseline under the same class and fight settings. A qualifying optic replaces iron sights when magnification is required.

New to Loadout Lab? Follow the six-step Quick Start or look up any option by its reference label.

Quick Start →Illustrated option guide
01 — DEFINE YOUR FIGHT

Your priorities. Your rules.

Advanced priorities & limits

Training effects are not numerically applied.

Time to kill

Firing scenario

Semi-auto assumes the maximum source cadence. Burst timing includes its source pauses. Ideal firing TTK excludes ADS and trigger-input delay.

Your priorities

Balanced defaults emphasize control, precision and lethality. Range and playstyle adjust their shares. ADS is one part of handling.

Visual recoil is a source ordinal preference. Its camera displacement is unverified.

Your limits

Ready. Click Find my builds to search and view Recommendations.

Find my builds takes you to Recommendations. Changing settings does not start a search. Exact can take several minutes; its progress appears there and beside this button.

02 — THE ARMORY

Build your kit

Choose a weapon

Tap a node to explore parts

Counts the parts equipped in this preview. Prefer and Exclude only guide the search.

FIG. 01 — ILLUSTRATIVE SCHEMATICAMBER = ACTIVE / CHANGED

PERFORMANCE MATRIXCurrent preview versus weapon baseline

Every change describes the current preview compared with this weapon’s baseline. Less recoil, spread or time is favorable. Ghost bars show baseline values, not your saved comparison build. Which reference am I reading?

Chart references & evidence

The radar is a display-only profile: damage ÷ 50, velocity ÷ 1000 m/s, fire rate ÷ 1000 RPM, and 1 − existing normalized recoil, precision and handling losses, capped at 100%. It never sets search rankings. The matrix shows midpoint values for range bands; recommendations evaluate the full sampled band.

The schematic is illustrative, with approximate attachment categories. It is not an exact weapon model or dimensional drawing. Native camera movement, armored REDSEC damage and some class effects still need evidence.

03 — ASK THE ARMORY

What's the best loadout for…

Name a supported weapon, class, distance or priority. Your request becomes search settings for the same optimizer.

Supported-phrase input, not a general chat assistant. Review the interpreted settings below. No invented attachment stats or preset builds.

OPTIMIZER TERMINAL
Awaiting request. Supported: weapon names, soldier class, mode, distance bands, recoil/handling/kill-time priorities, optic magnification and laser preference.
View search results ↓

THREE WAYS TO FIT YOUR FIGHT

Your recommendations

Read these results · J1

Choose your fight and find your builds.

Choose attachments · Must use, Prefer or Exclude

Every source-catalog attachment stays available. Must use locks the exact part and cost. Prefer adds a small, disclosed preference penalty to other choices. Exclude removes a part from the search.

First-time exact searches can take minutes. Completed matching jobs are reused.
Choose your reference build for comparison

Your reference is the build you already use or want to compare against. Search results remain the recommendations. Changes describe the recommendation relative to this reference. Comparison guide · K5

Enter what you already use. Comparisons rebuild both weapons under the same current class, range and firing assumptions. A current build may fall outside your search restrictions; those differences will be shown.

YOUR OWN ARMORY

Saved builds

Saved on this browser. Share links include the build, settings and data version; the recipient still needs access to this site. Data/model changes flag older builds for review.

OPEN NOTEBOOK

What the numbers mean

Complete choices, exact verification

Weapon-specific optics, magazines, ammo, muzzles, barrels, grips, lasers, lights, shared rails, ergonomics and accessories are available wherever the source catalog permits them. Shared rails hold one device. Dependency rules reject incompatible combinations, including magazine/grip conflicts.

Exact mode accounts for the complete raw choice space, enumerating effect-distinct builds or safely pruning branches. Equivalence preserves all attachment effect data and dependency roles. Minimum remaining-cost bounds reject over-budget branches; audited lower score bounds reject branches that cannot improve the retained balance, control or kill-timing objectives. Locked attachments are applied before equivalence and keep their exact cost. The completion receipt accounts for the raw choice space. Fast mode remains an optional beam-search shortlist.

Operator effects

Assault AR sprint/draw, Support LMG ADS, Engineer SMG hip spread and Recon sniper sway are conditionally modeled. Each card shows the bonus separately from its attachment effects. Signature-family filtering is on by default; turn it off for open-weapon comparisons.

Patch 2.1.1.0 source research rechecks the class operands and confirms Mk 24, Z84M, RSPR and ELR-40 bindings (L4070). Native activation and composition remain conditional; this is not an independent in-game test. DMRs have no signature-weapon bonus. Training-path effects, Recon rechamber/breath bonuses and Support sprint-speed exemption are not numerically modeled.

Perfect hits, with travel shown

All shots hit the selected zone. Firing TTK starts at shot one and follows damage breakpoints and firing cadence. Estimated impact TTK adds range ÷ initial bullet speed once. That travel estimate ignores drag and trajectory; it is not a validated ballistic simulation. No aim or sprint delay is added.

Armored REDSEC TTK remains unknown. An explicitly unarmored reference is available. Builds needing a reload to kill are not assigned a TTK. Sniper timing uses the source baseline cycle; the unquantified Recon cycle bonus is excluded.

Sources, remaining gaps and version

Data snapshot: 2.1.1.0, published source reviewed 8 Nov 2026. Version 0.8.0 supports 56 primary weapons, including all six source DMRs and M16A4. VSSM can also be found as VSS. Shotguns and sidearms are not optimized. Exports retain the exact source commit, data fingerprint and model hashes.

Recovery and sustained spread now participate through an assumed firing simulation. Laser visibility and utility requirements constrain builds, with optional minimap-exposure and ordinal visual preferences. Season 5 class bindings are source-backed by L4070; native class activation and perk effects remain conditional. Armor curves and materials have partial evidence (L4076), but EA documents 40 armor HP per plate and 80 total in Battle Royale. Armor curve selection, modifier composition and the amount of damage transferred into health on an armor-breaking hit remain unresolved. Camera motion, backup/variable optics, world spotting and the complete native physics equations still need qualified evidence. Each result includes a coverage table. EA armor capacity · EA armor-breaking damage events.

Numerical source ↗ · Class evidence ↗ · EA classes ↗ · Versioned dataset ↓
Faster exact search

Every catalog attachment stays selectable. Lock a named sight to reserve its exact point cost and build around it. Exact search uses a strong provisional starting build, safe cost/score bounds and up to four background workers. Recoil and spread calculations are reused. Completed search jobs are saved on this browser for the same settings and data/model version; changing priorities, class, range or attachment locks triggers a fresh search. Pause or stop at any time. Broad first-time searches can still take substantial time. Only a completed search certifies the best result for the modeled stats.

Scoring formula