---
title: "Compare calibers"
method: GET
path: "/v1/calibers/compare"
tags: ["Calibers"]
---

# Compare calibers

`GET /v1/calibers/compare`

## Request

- Base URL: `https://api.gunspec.io`
- URL: `https://api.gunspec.io/v1/calibers/compare`
- Auth: one of:
  - API key in header `X-API-Key`
  - HTTP bearer

Returns full specifications for up to 5 cartridges side by side.

**Requires Builder tier or higher.**

## Query parameters

- `ids` string, required

## Headers

- `If-None-Match` string

## Response `200`

Compare calibers

- object
  - `success` boolean, required: Always true on a successful response. Every error carries `false` and an `error` object instead.
  - `data` object, required: The response payload.
    - `items` Caliber[]: The compared cartridges, in the order their ids were given.
      - `id` string: URL-safe slug identifying the cartridge, e.g. `9x19mm-parabellum`.
      - `name` string: Display name, as the maker writes it.
      - `aliases` string[], nullable: Other names this cartridge is sold or written as.
      - `natoDesignation` string, nullable: NATO designation where the cartridge has one, e.g. `9mm NATO`. Null otherwise.
      - `bulletDiameterMm` number, nullable: Bullet diameter in millimetres.
      - `neckDiameterMm` number, nullable: Case neck outside diameter in millimetres.
      - `shoulderDiameterMm` number, nullable: Bottleneck cases only.
      - `baseDiameterMm` number, nullable: Case base diameter in millimetres, ahead of the extractor groove.
      - `rimDiameterMm` number, nullable: Rim diameter in millimetres.
      - `rimThicknessMm` number, nullable: Rim thickness in millimetres.
      - `caseLengthMm` number, nullable: Case length in millimetres.
      - `overallLengthMm` number, nullable: Overall length in millimetres, stock extended where it folds or collapses.
      - `bulletLengthMm` number, nullable: Typical projectile length.
      - `maxPressureMpa` number, nullable: Maximum average pressure in megapascals, as the standards body specifies it.
      - `maxPressurePsi` number, nullable: The same pressure in pounds per square inch, converted.
      - `typicalBulletWeightG` number, nullable: Representative bullet weight in grams. Divide by 0.0648 for grains.
      - `typicalMuzzleVelocityMps` number, nullable: Representative muzzle velocity in metres per second.
      - `typicalMuzzleEnergyJ` number, nullable: Representative muzzle energy in joules.
      - `primerType` string, nullable: Primer size and type, e.g. `small_pistol`, `large_rifle`, `rimfire`.
      - `cartridgeType` string, nullable: Head type: `rimmed`, `rimless`, `semi_rimmed`, `rebated_rim`, `belted`, plus historical values like `cap_and_ball`. **Not the broad class.** The registry said `pistol`/`rifle`/`shotgun` until the column was read, and it holds neither.
      - `caseShape` 'straight' | 'tapered' | 'bottleneck' | 'none' | 'null', nullable: Case profile: `straight`, `bottleneck`, `tapered`, `belted`.
      - `caseMaterial` 'brass' | 'nickel_brass' | 'steel' | 'lacquered_steel' | 'aluminium' | 'polymer' | 'paper' | 'caseless' | 'none' | 'null', nullable: What the case is made of: `brass`, `steel`, `aluminium`, `polymer`.
      - `projectileKind` 'bullet' | 'none' | 'shot' | 'slug' | 'round_ball' | 'flechette' | 'dart' | 'pellet' | 'rocket' | 'grenade' | 'signal' | 'null', nullable: `none` is a blank.
      - `bulletProfile` 'spitzer' | 'spitzer_boat_tail' | 'round_nose' | 'flat_nose' | 'wadcutter' | 'semi_wadcutter' | 'hollow_point' | 'truncated_cone' | 'round_ball' | 'null', nullable: Bullet shape: `fmj`, `hollow_point`, `soft_point`, `spitzer`, `round_nose`.
      - `closure` 'bullet' | 'star_crimp' | 'rosette_crimp' | 'roll_crimp' | 'fold_crimp' | 'wad' | 'paper' | 'none' | 'null', nullable: What seals the case mouth; a blank is a star crimp.
      - `markingColor` 'black' | 'white' | 'red' | 'orange' | 'yellow' | 'green' | 'blue' | 'purple' | 'pink' | 'brown' | 'grey' | 'silver' | 'gold' | 'null', nullable: Paint that identifies the load: crimp lacquer on a blank, tip colour on a military round.
      - `markingMeaning` string, nullable: What that colour signifies, where the marking is a standard one.
      - `parentCartridgeId` string, nullable: The cartridge this one was derived from, or null. A wildcat and its parent case, for example.
      - `yearIntroduced` integer, nullable: Year it entered production or service. Null where the date is disputed or unknown.
      - `designer` string, nullable: Person or team credited with the design, where it is known.
      - `specSource` string, nullable: Page the dimensions were read from.
      - `specStandard` 'saami' | 'cip' | 'nato' | 'proprietary' | 'none' | 'null', nullable: The standards body whose figures these are (`SAAMI`, `CIP`), where the source names one.
      - `dataConfidence` number, nullable: How completely the record is specified and how well it is sourced, 0 to 1. A record-level completeness and provenance measure, **not** a per-field probability of correctness. Use it to rank and triage; use `sources` to verify an individual number.
      - `verifiedAt` string, nullable: When a person last checked this record against its source. Null means nobody has: the row is model knowledge until then.
      - `verifiedFields` string[], nullable: The fields the source stated. The same list is in `provenance.verifiedFields`.
      - `createdAt` string: When the record was first added, `YYYY-MM-DD HH:MM:SS` in UTC.
      - `updatedAt` string: When a field we serve last changed. Maintained by the database rather than by whatever wrote the row, so it moves for a corrected specification and does not move for a re-run of the importer.
      - `version` string, nullable: Opaque fingerprint of this record. It changes when and only when a field we serve changes, so a client holding the same value already has the record and need not fetch it. Equality is the only supported operation: do not parse it, and do not assume how it is computed. Null means the record has not been written since versioning began, not that it is unchanged.
      - `provenance` Provenance: Where a record's figures came from and how far they have been checked, in one place. `sources` are the pages consulted; check a specific figure against those. `sourceKinds` says what each page is and `bestSourceKind` the strongest of them, on the hierarchy `SourceKind` defines: a maker or standards-body page is evidence for a figure, a retailer or forum page is evidence the item exists. `dataConfidence` is the 0 to 1 score set from what was actually sourced and never raised by hand. `verifiedAt` and `verifiedFields` say when a source was last read against the record and which fields it stated; null means the row is still seed knowledge. `updatedAt` and `version` are the same cache signals the record carries at the top level.
        - `sources` string[], required: The pages consulted when the record was compiled, as an array. Check a specific figure against these rather than against `dataConfidence`.
        - `sourceKinds` object[], required: What each cited page is, in `sources` order, on the `SourceKind` hierarchy. A host not in the source map is `other`, never guessed.
          - `url` string, required: Absolute, directly fetchable URL.
          - `kind` 'manufacturer' | 'standards_body' | 'government' | 'peer_reviewed' | 'reference' | 'aggregator' | 'press' | 'retailer' | 'community' | 'other', required: What kind of page a cited source is, strongest first: the maker's own page, a standards body (SAAMI, C.I.P.), a government or military document, an independent reference work, a specification aggregator, the press, a retailer or auction listing, a community site, or an unclassified page. Weigh a figure by the best kind behind it, never by how many pages were cited.
          - `archivedFrom` string, nullable, required: The original address when `url` is a Wayback Machine capture of it. A cited page that no longer answers is replaced by its last capture that answered 200, so the citation still shows what its author read; `kind` is the original page's.
          - `rank` integer, required: The kind's place in the source hierarchy: 0 is the maker, the strongest; higher is weaker.
          - `tier` 'authority' | 'classified' | 'unclassified', required: Where the site stands in the source registry (`/v1/data/sources`): an authority (the maker, a standards body, a government), a classified site, or one not yet classed.
          - `basis` 'declared' | 'maker' | 'rule' | 'unclassified', required: What the class rests on: classed by a person, a catalogued maker's own site, a .gov or .mil domain, or nothing yet.
          - `evidence` string, nullable, required: The words the site was classed on, where a person recorded them.
        - `bestSourceKind` 'manufacturer' | 'standards_body' | 'government' | 'peer_reviewed' | 'reference' | 'aggregator' | 'press' | 'retailer' | 'community' | 'other', required: What kind of page a cited source is, strongest first: the maker's own page, a standards body (SAAMI, C.I.P.), a government or military document, an independent reference work, a specification aggregator, the press, a retailer or auction listing, a community site, or an unclassified page. Weigh a figure by the best kind behind it, never by how many pages were cited.
        - `makerSource` string, nullable, required: The first cited page on the record's own manufacturer's website, or null when it cites none. A page on another maker's site does not count. Cited from the maker is not the same as checked against it: `verifiedFields` names the figures that were read against a source.
        - `evidence` 'verified' | 'maker' | 'authority' | 'secondary' | 'none', required: How far the record's figures are backed, strongest first, for a badge: `verified` when at least one figure was read against a source (`verifiedFields`), `maker` when it cites a page on its own manufacturer's site (`makerSource`), `authority` when it cites a maker, standards body or government page, `secondary` when it cites only press, retailers, references or forums, `none` when it cites nothing. Derived from the fields beside it.
        - `checks` object: The checks the record has been through, on a single record's response only: every task the data agent (Redback) or a person was given about it, what each proposed, the pages it read, and whether a person accepted it. Nothing an agent proposes reaches the record until a person accepts it, so this is the trail behind `verifiedFields`. Who reviewed is not published; an agent's instance name is.
          - `raised` integer, required: Every task ever raised on the record, answered or not.
          - `open` integer, required: Tasks still waiting to be worked.
          - `history` object[], required: The latest twenty, newest first.
            - `task` string, required: The task id, stable across runs; the same id `/v1/data/tasks/{taskKey}` answers for.
            - `number` integer, nullable, required: The task's number; `/v1/data/tasks/{number}` loads it as its key does.
            - `check` string, nullable, required: The check that raised it, where one did.
            - `kind` string, required: `verify` reads the record against its maker's page, `gap` researches an empty field, `contradiction` judges fields that disagree, `vocabulary` a value outside a list.
            - `agent` string, nullable, required: `redback` when the agent worked it, `staff` when a person did, null when nobody has.
            - `worker` string, nullable, required: The agent instance that worked it; null for a person.
            - `status` string, required: `resolved` (a proposal was filed), `unverifiable` (nothing could be established; retried after a cooldown), `not_a_fault`, or open.
            - `decidedAt` string, nullable, required: When the task was answered; null while it is open.
            - `proposed` string, nullable, required: What it proposed, in its own words; null while it is open.
            - `fields` string[], required: Per-field completeness, keyed by column name.
            - `sources` string[], required: The pages it read and quoted.
            - `review` object, nullable, required: A person's decision on the proposal; null until reviewed.
              - …
            - `finding` object, nullable, required: What the check found, where a check raised the task: its own line for this record and the figures it measured, so a reader sees how the value breaks the method. The same the public task board serves with `include=research`.
              - …
        - `dataConfidence` number, nullable, required: The 0 to 1 score set from what was actually sourced, never raised by hand. See the confidence bands in the docs.
        - `verifiedAt` string, nullable, required: When a source was last read against this record. Null means the row is still seed knowledge.
        - `verifiedFields` string[], nullable, required: Which fields the source stated, as an array. Everything else on a verified row is still unverified. Null where no source has been read.
        - `specSource` string, nullable, required: Cartridges only: the page the drawing figures were taken from. Null on every other kind of record.
        - `updatedAt` string, required: When a served column last changed; the same value the record carries at the top level.
        - `version` string, nullable, required: The record's content version, equal to the top-level `version`; equal versions mean equal data.
    - `deltas` object[]: Per-field differences across the compared cartridges, one entry per numeric field.
      - `field` string: Which column the row is about, named as the database names it.
      - `values` number[]: This field on each compared cartridge, in the order the ids were given. Null where that record does not carry the field.
      - `min` number, nullable: Lowest value across the compared cartridges. Null when fewer than two carry the field.
      - `max` number, nullable: Highest value across the compared cartridges. Null when fewer than two carry the field.
      - `percentDiff` number, nullable: Spread from the lowest value, as a percentage of it. Null when the minimum is zero or fewer than two carry the field.

## Other responses

- `304`: Your copy is current. No body is sent, and this response does not count toward the daily request allowance.
- `400`: Validation error
- `401`: Missing or invalid API key
- `403`: The credential is valid and the caller is not permitted. Three causes, none of which reissuing a key changes: the plan is below the endpoint's tier, the account is suspended, or the key is not attached to the thing being acted on, meaning a user account for `/v1/me` or a shop for `/v1/vendor`. Distinct from 401, which means the credential is missing, malformed or expired and a new one can work.
- `429`: A limit was reached. `RATE_LIMITED` is the per-minute limit: wait `Retry-After` seconds (a minute) and retry. `DAILY_CAP_EXCEEDED` is the plan's daily limit, per key: nothing will be served until the counters return to zero at midnight UTC. `MONTHLY_CAP_EXCEEDED` is the plan's monthly allowance, for the whole account: nothing will be served until midnight UTC on the first of next month, or until the plan changes. For both, `Retry-After` is the seconds until then and `error.resetsAt` says when as an instant. Calls made after a refusal are refused too and do not spend the month, and a free key that keeps calling is paused, so stop at the first one and resume at the reset.
- `500`: Internal server error

---

[API](https://skmtc.dev/gunspec/apis/gunspec-io-api.md) · [All operations](https://skmtc.dev/gunspec/apis/gunspec-io-api/llms.txt) · [OpenAPI document](https://skmtc.dev/gunspec/apis/gunspec-io-api/revisions/f2e974d8b05c?raw)