Every successful call earns experience for the account that made it, weighted by what the call was. The intent is to recognise how deeply you have integrated rather than how much traffic you send, so the calls you only make once you have built something are worth several times a page of a list.
Read from the table the API writes from, so this page cannot fall behind it. 20 operations are worth more than the floor; everything else earns 1.
| Operation | What it is | Worth |
|---|---|---|
GET /v1/attachments/{id}/firearms | List firearms an attachment fits | 8 |
GET /v1/firearms/{id}/attachments | List attachments that fit a firearm | 8 |
GET /v1/firearms/{id}/interfaces | Get a firearm's mount interfaces | 8 |
GET /v1/firearms/head-to-head | Compare two firearms head to head | 8 |
GET /v1/firearms/resolve | Resolve a name to one firearm | 8 |
POST /v1/firearms/resolve | Resolve many names at once | 8 |
GET /v1/ammunition/{id}/ballistics | Compute a ballistic table | 5 |
GET /v1/calibers/{id}/firearms | List firearms in a caliber | 5 |
GET /v1/calibers/ballistics | Get a caliber ballistic profile | 5 |
GET /v1/calibers/compare | Compare calibers | 5 |
GET /v1/firearms/{id}/calculate | Calculate ballistics for a firearm | 5 |
GET /v1/firearms/{id}/similar | Find similar firearms | 5 |
GET /v1/firearms/{id}/variants | List variants | 5 |
GET /v1/firearms/compare | Compare firearms | 5 |
GET /v1/firearms/search | Full-text search | 3 |
GET /v1/ammunition/{id} | Get an ammunition load | 2 |
GET /v1/attachments/{id} | Get an attachment | 2 |
GET /v1/calibers/{id} | Get a caliber | 2 |
GET /v1/firearms/{id} | Get a firearm | 2 |
GET /v1/manufacturers/{id} | Get a manufacturer | 2 |
| Every other operation | A page of a list included. Paging is how a mirror is built, which is fine, and not a demonstration of anything. | 1 |
Experience recognises something you built, so it is earned by the clients that do the building. The kind is read from your User-Agent, the same way the request log records it.
| Client | Earns | What it is |
|---|---|---|
sdk-typescript | yes | The TypeScript SDK. |
sdk-python | yes | The Python SDK. |
mcp | yes | The hosted MCP server. |
http-tool | yes | curl, Postman, HTTPie and the like: somebody trying it for real. |
http-library | yes | An HTTP client inside your own code. |
unknown | yes | A runtime that sends no agent. Far likelier to be your integration than a browser. |
browser | no | A browser earns nothing. |
Two other calls earn nothing. A 304 Not Modified earns nothing, on the same reasoning that it spends none of your allowance: the polite integration must not be the expensive one, and caching is what we would rather you did. A call that failed earns nothing either, so guessing at ids pays nothing.
Your level follows from your total and nothing else. Each level costs a little more than the last, so the early ones come quickly and the later ones are earned. The days column is what it would take at the ceiling of 500 a day, which a real integration never meets.
| Level | Starts at | Costs | Days at the cap |
|---|---|---|---|
| 1 | 0 | 0 | 0 |
| 2 | 50 | 50 | 1 |
| 3 | 200 | 150 | 1 |
| 4 | 450 | 250 | 1 |
| 5 | 800 | 350 | 2 |
| 6 | 1,250 | 450 | 3 |
| 7 | 1,800 | 550 | 4 |
| 8 | 2,450 | 650 | 5 |
| 9 | 3,200 | 750 | 7 |
| 10 | 4,050 | 850 | 9 |
Two figures beside the level, both counted rather than earned. There are 145 documented operations to meet.
operationsUsed is how many you have called at least once, against operationsAvailable. Counted by a nightly job from the request log, so an operation you call for the first time today appears tomorrow. It costs nothing on the request path, which is the point.
streakDays is consecutive days you have called at all, and daysActive is the total. Yesterday still counts as unbroken, so a timezone does not cost you a streak. Neither pays experience: a bonus for showing up rewards the calendar rather than the integration.
GET /v1/me/usage returns a progress block beside the usage counters: xp, xpToday, dailyXpCap, level, levelStartsAt, nextLevelAt, and the breadth and streak figures above. The two level bounds are there so a client can draw the bar without knowing the formula.
| Field | What it means |
|---|---|
xp | Experience earned across every key on the account, for its whole life. |
xpToday | Experience earned today (UTC). |
dailyXpCap | The most one key can earn in a UTC day. A ceiling rather than a target: reaching it every day is not how the levels are meant to be climbed. |
level | Your level, derived from xp alone. Nothing else moves it. |
levelStartsAt | The experience this level began at. |
nextLevelAt | The experience the next level begins at. |
streakDays | Consecutive days you have called the API, counting today. A day you did not call breaks it, though yesterday still counts as unbroken so a timezone does not cost you a streak. Reported, never rewarded: no experience is paid for showing up. |
daysActive | Days you have called the API at all, over the last two years. |
operationsUsed | Documented operations you have called at least once. Counted from the request log by the nightly job, so an operation you called for the first time today appears tomorrow. |
operationsAvailable | Documented operations there are to call, counted from this document rather than written down, so it cannot fall behind a route added today. |
