Skip to content
Sign up

flagon-io/g1t

public
0
1,103 linesCodeBlameRaw

Pick any line to see why it is the way it is: the commit, the pull request and issue it came from, and what the agent was thinking.

Prices keep themselves current with what g1t pays1//! Keeps every price current with what g1t actually pays.
2//!
3//! g1t passes its own costs through, so a price is only right while the
4//! cost under it is. Two jobs keep them right, on the billing service's
5//! cron:
6//!
7//! - **Settling runs** (every 15 minutes). A model run is charged when it
8//! finishes at what the sandbox reported. Each of g1t's hosted runs goes
9//! through its AI Gateway, which prices every request at the provider's
10//! current rates and logs it with the run's session. Settling sums those
11//! logs and corrects the charge to the gateway's figure, with a
12//! correction on the statement. A run whose sandbox died before
13//! reporting is charged here instead of never.
14//! - **Checking costs** (daily). What Cloudflare billed g1t's account, from
15//! its usage API, is measured against how much was used: Containers
16//! against the seconds containers ran, Workers for Platforms per request
17//! and per CPU millisecond. When a measured cost moves, the price book
18//! moves with it, since each price is its cost plus a set markup, and
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily19//! the change is recorded where anyone can see it. A measurement goes
20//! through `pricing` as a proposal: a small move is applied on its own (a
21//! rise only after customers have had notice), a large or suspect one
22//! waits for staff in sudo, so one odd day of data cannot reprice
23//! everything.
Prices keep themselves current with what g1t pays24
25use g1t_contracts::billing::{EntryKind, MICROS_PER_DOLLAR, Price, PriceBook, PriceChange};
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily26
Prices keep themselves current with what g1t pays27use g1t_contracts::time::rfc3339;
28use g1t_kit::now_ms;
29use serde::Deserialize;
30use serde_json::{Value, json};
31use worker::{Env, Fetch, Headers, Method, Request, RequestInit, Result};
32
Merge branch 'worktree-agent-a633ac0f7f66d419d'33use crate::{Billing, RunRow};
Prices keep themselves current with what g1t pays34
35/// The cron that also checks costs against Cloudflare's bill.
36pub(crate) const DAILY: &str = "17 4 * * *";
Merge platform pause and the hourly usage watcher: staff can pause compute, schedules, indexing or renders for everyone, the watcher emails on a breach and is never blind quietly, and the models proxy holds each run to its cap (billing 0051, integrations 0006)37/// The quarter-hourly tick: settling, and once an hour the platform watch.
38pub(crate) const QUARTER_HOURLY: &str = "*/15 * * * *";
Prices keep themselves current with what g1t pays39
40/// A run is settled once its logs have had time to land.
41const SETTLE_AFTER_MS: u64 = 5 * 60 * 1000;
42/// A run with no gateway logs after this is left as reported.
43const GIVE_UP_AFTER_MS: u64 = 3 * 60 * 60 * 1000;
44/// A run never finished after this died without reporting.
45const ABANDONED_AFTER_MS: u64 = 3 * 60 * 60 * 1000;
46
47/// Where the keeper reads what g1t pays.
48pub(crate) struct Keeper {
49 /// `CLOUDFLARE_USAGE_TOKEN`: Billing, Account Analytics and AI Gateway,
50 /// read only.
51 token: Option<String>,
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily52 /// What reads the bill for `costs`: `CLOUDFLARE_BILLING_TOKEN`
53 /// (Account: Billing Read and Account Analytics Read), or the usage
54 /// token, which has both.
55 billing_token: Option<String>,
Prices keep themselves current with what g1t pays56 account: String,
57 gateway: String,
58}
59
60impl Keeper {
61 pub(crate) fn from_env(env: &Env) -> Self {
62 let var = |name: &str| env.var(name).map(|v| v.to_string()).unwrap_or_default();
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily63 let secret = |name: &str| env.secret(name).ok().map(|v| v.to_string()).filter(|v| !v.is_empty());
64 let token = secret("CLOUDFLARE_USAGE_TOKEN");
Prices keep themselves current with what g1t pays65 Keeper {
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily66 billing_token: secret("CLOUDFLARE_BILLING_TOKEN").or_else(|| token.clone()),
67 token,
Prices keep themselves current with what g1t pays68 account: var("CLOUDFLARE_ACCOUNT_ID"),
69 gateway: var("AI_GATEWAY_ID"),
70 }
71 }
72
73 async fn send(&self, method: Method, url: &str, body: Option<Value>) -> Result<Value> {
74 let Some(token) = &self.token else {
75 return Err(worker::Error::RustError("no CLOUDFLARE_USAGE_TOKEN".into()));
76 };
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily77 send_with(token, method, url, body).await
78 }
79
80 /// Whether Cloudflare's bill can be read.
81 pub(crate) fn can_read_bill(&self) -> bool {
82 self.billing_token.is_some() && !self.account.is_empty()
83 }
84
85 fn billing_token(&self) -> Result<&str> {
86 self.billing_token
87 .as_deref()
88 .ok_or_else(|| worker::Error::RustError("no CLOUDFLARE_BILLING_TOKEN or CLOUDFLARE_USAGE_TOKEN".into()))
89 }
90
Merge Cloudflare's usage over its billing cycle: every page read, included amounts once a cycle, a projection, test-mode charges never money in (billing 0052)91 /// Billable usage from `from` to `to` (dates), every page of it, as one
92 /// answer (`result` holds all the rows), and how many pages it took.
93 /// An answer that says it failed is returned as it is.
94 pub(crate) async fn billable_usage_pages(&self, from: &str, to: &str) -> Result<(Value, u32)> {
95 usage_pages(self.billing_token()?, &self.api(&format!("/billable-usage?from={from}&to={to}"))).await
Prices keep themselves current with what g1t pays96 }
97
Costs: Cloudflare's subscriptions read from Cloudflare each day, the estimate only until then; sudo's costs split into Costs & margin and Bill & pricing98 /// The account's subscriptions (Workers Paid, add-ons), as Cloudflare
99 /// answers them. Needs Account: Billing Read.
100 pub(crate) async fn subscriptions_body(&self) -> Result<Value> {
101 send_with(self.billing_token()?, Method::Get, &self.api("/subscriptions"), None).await
102 }
103
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily104 /// A GraphQL Analytics query, as Cloudflare answers it, errors and all.
105 pub(crate) async fn graphql(&self, body: Value) -> Result<Value> {
106 send_with(self.billing_token()?, Method::Post, "https://api.cloudflare.com/client/v4/graphql", Some(body)).await
107 }
108
109 pub(crate) fn account(&self) -> &str {
110 &self.account
111 }
112
Prices keep themselves current with what g1t pays113 fn api(&self, path: &str) -> String {
114 format!("https://api.cloudflare.com/client/v4/accounts/{}{path}", self.account)
115 }
116
Merge branch 'worktree-agent-a633ac0f7f66d419d'117 /// AI Gateway's id, empty when there is none.
118 pub(crate) fn gateway(&self) -> &str {
119 &self.gateway
120 }
121
Billing: AI Gateway's analytics are read with the token that can see them, and a gateway that priced nothing is said122 /// A GraphQL query over AI Gateway's analytics: with the keeper's token
Merge costs and margin review: gateway query, own spend, discount meters, superseded rises123 /// (AI Gateway Read) first, and on failure with the bill's. A token
124 /// without Account Analytics Read gets an error ("caller does not hold
125 /// any of the required permissions for this dataset"); when the answer
126 /// has no rows, `gateway_visible` tells a token that cannot see the
127 /// gateway from a gateway nothing went through.
Billing: AI Gateway's analytics are read with the token that can see them, and a gateway that priced nothing is said128 pub(crate) async fn gateway_graphql(&self, body: Value) -> Result<Value> {
129 let Some(token) = &self.token else {
130 return self.graphql(body).await;
131 };
132 match send_with(token, Method::Post, "https://api.cloudflare.com/client/v4/graphql", Some(body.clone())).await {
Merge branch 'worktree-agent-a633ac0f7f66d419d'133 Ok(answer) => Ok(answer),
Billing: AI Gateway's analytics are read with the token that can see them, and a gateway that priced nothing is said134 Err(error) => match &self.billing_token {
135 Some(billing) if billing != token => self.graphql(body).await,
Merge branch 'worktree-agent-a633ac0f7f66d419d'136 _ => Err(error),
137 },
138 }
139 }
140
Merge costs and margin review: gateway query, own spend, discount meters, superseded rises141 /// Whether the token AI Gateway's analytics are read with can see the
142 /// gateway: the REST API answers 403 for a token without AI Gateway
143 /// Read and 404 for a gateway id that is not there. None when the
144 /// answer says neither (Cloudflare down, no token).
145 pub(crate) async fn gateway_visible(&self) -> Option<bool> {
146 let token = self.token.as_deref().or(self.billing_token.as_deref())?;
147 match send_with(token, Method::Get, &self.api(&format!("/ai-gateway/gateways/{}", self.gateway)), None).await {
148 Ok(_) => Some(true),
149 Err(error) => refused(&error.to_string()).then_some(false),
150 }
151 }
152
Merge branch 'worktree-agent-a633ac0f7f66d419d'153 /// What AI Gateway priced a session's requests at, and how many there
154 /// were, with the requests it had no price for.
155 async fn session_cost(&self, session: &str) -> Result<SessionCost> {
156 let mut total = SessionCost { complete: true, ..SessionCost::default() };
157 for page in 1..=MAX_LOG_PAGES {
The keeper reads Cloudflare as it really answers158 // The filter goes as URL-encoded JSON; the bracket form is
159 // ignored, and would sum every log there is. Session ids are
160 // [a-z0-9_], which need no escaping inside it.
161 let filter = format!(
162 "%5B%7B%22key%22%3A%22metadata.value%22%2C%22operator%22%3A%22eq%22%2C%22value%22%3A%5B%22{session}%22%5D%7D%5D"
163 );
Prices keep themselves current with what g1t pays164 let url = self.api(&format!(
The keeper reads Cloudflare as it really answers165 "/ai-gateway/gateways/{}/logs?per_page=50&page={page}&filters={filter}",
Prices keep themselves current with what g1t pays166 self.gateway
167 ));
168 let body = self.send(Method::Get, &url, None).await?;
169 let logs = body["result"].as_array().cloned().unwrap_or_default();
170 for log in &logs {
Merge branch 'worktree-agent-a633ac0f7f66d419d'171 total.add(log);
Prices keep themselves current with what g1t pays172 }
173 if logs.len() < 50 {
Merge branch 'worktree-agent-a633ac0f7f66d419d'174 return Ok(total);
Prices keep themselves current with what g1t pays175 }
176 }
Merge branch 'worktree-agent-a633ac0f7f66d419d'177 // More logs than were read: what was read is less than the run.
178 total.complete = false;
179 Ok(total)
Prices keep themselves current with what g1t pays180 }
181
The keeper reads Cloudflare as it really answers182 /// The account's billable usage, one row per service per day, as
183 /// Cloudflare reports it.
Prices keep themselves current with what g1t pays184 async fn billable_usage(&self, from: &str, to: &str) -> Result<Vec<UsageRow>> {
Merge Cloudflare's usage over its billing cycle: every page read, included amounts once a cycle, a projection, test-mode charges never money in (billing 0052)185 let Some(token) = &self.token else {
186 return Err(worker::Error::RustError("no CLOUDFLARE_USAGE_TOKEN".into()));
187 };
188 let (body, _) = usage_pages(token, &self.api(&format!("/billable-usage?from={from}&to={to}"))).await?;
Prices keep themselves current with what g1t pays189 let rows = body["result"].as_array().cloned().unwrap_or_default();
190 Ok(rows.iter().filter_map(UsageRow::from_value).collect())
191 }
192
The keeper reads Cloudflare as it really answers193 /// What g1t's containers used from `since` to `until` (dates), as
194 /// Cloudflare bills it: memory in byte-seconds, and CPU seconds.
195 async fn container_usage(&self, since: &str, until: &str) -> Result<ContainerUsage> {
196 let query = "query ($account: String!, $since: Date!, $until: Date!) {
Prices keep themselves current with what g1t pays197 viewer { accounts(filter: { accountTag: $account }) {
The keeper reads Cloudflare as it really answers198 containersUsageAdaptiveGroups(limit: 1000, filter: { date_geq: $since, date_leq: $until }) {
199 sum { cpuTimeSec allocatedMemory }
Prices keep themselves current with what g1t pays200 }
201 } }
202 }";
203 let body = self
204 .send(
205 Method::Post,
206 "https://api.cloudflare.com/client/v4/graphql",
207 Some(json!({ "query": query, "variables": { "account": self.account, "since": since, "until": until } })),
208 )
209 .await?;
The keeper reads Cloudflare as it really answers210 let groups = body["data"]["viewer"]["accounts"][0]["containersUsageAdaptiveGroups"]
Prices keep themselves current with what g1t pays211 .as_array()
212 .cloned()
213 .unwrap_or_default();
The keeper reads Cloudflare as it really answers214 Ok(groups.iter().fold(ContainerUsage::default(), |total, g| ContainerUsage {
215 cpu_seconds: total.cpu_seconds + g["sum"]["cpuTimeSec"].as_f64().unwrap_or(0.0),
216 memory_byte_seconds: total.memory_byte_seconds + g["sum"]["allocatedMemory"].as_f64().unwrap_or(0.0),
217 }))
Prices keep themselves current with what g1t pays218 }
219}
220
Merge costs and margin review: gateway query, own spend, discount meters, superseded rises221/// Whether an error from `send_with` is Cloudflare saying no to the token
222/// (401, 403) or that there is no such thing for it (404), rather than
223/// failing.
224pub(crate) fn refused(error: &str) -> bool {
225 ["Cloudflare answered 401", "Cloudflare answered 403", "Cloudflare answered 404"].iter().any(|s| error.contains(s))
226}
227
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily228/// A request to Cloudflare's API with a bearer token; anything but 200 is
229/// an error with what Cloudflare said.
Merge Cloudflare's usage over its billing cycle: every page read, included amounts once a cycle, a projection, test-mode charges never money in (billing 0052)230/// The most pages of billable usage read in one go: far more than a few
231/// months of g1t's meters.
232const MAX_USAGE_PAGES: u32 = 50;
233
234/// Every page of a billable-usage answer, its rows together. Before
235/// 2026-10-09 only the first page was read.
236async fn usage_pages(token: &str, url: &str) -> Result<(Value, u32)> {
237 let mut rows: Vec<Value> = Vec::new();
238 let mut pages = 0;
239 let mut next = url.to_owned();
240 loop {
241 let body = send_with(token, Method::Get, &next, None).await?;
242 if body["success"] == Value::Bool(false) {
243 return Ok((body, pages + 1));
244 }
245 pages += 1;
246 rows.extend(body["result"].as_array().cloned().unwrap_or_default());
247 match crate::costs::next_page(&body, pages).filter(|_| pages < MAX_USAGE_PAGES) {
248 Some(query) => next = format!("{url}&{query}"),
249 None => break,
250 }
251 }
252 Ok((json!({ "success": true, "result": rows }), pages))
253}
254
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily255async fn send_with(token: &str, method: Method, url: &str, body: Option<Value>) -> Result<Value> {
256 let headers = Headers::new();
257 headers.set("authorization", &format!("Bearer {token}"))?;
258 headers.set("content-type", "application/json")?;
259 let mut init = RequestInit::new();
260 init.with_method(method).with_headers(headers);
261 if let Some(body) = body {
262 init.with_body(Some(body.to_string().into()));
263 }
264 let mut response = Fetch::Request(Request::new_with_init(url, &init)?).send().await?;
265 let status = response.status_code();
266 let value: Value = response.json().await.unwrap_or(Value::Null);
267 if status != 200 {
268 return Err(worker::Error::RustError(format!("Cloudflare answered {status}: {value}")));
269 }
270 Ok(value)
271}
272
The keeper reads Cloudflare as it really answers273#[derive(Debug, Default, Clone, Copy)]
274pub(crate) struct ContainerUsage {
275 cpu_seconds: f64,
276 memory_byte_seconds: f64,
277}
278
279/// g1t's sandboxes: Containers' standard-1, half a vCPU, 4 GiB, 8 GB disk.
280const SANDBOX_GIB: f64 = 4.0;
281const SANDBOX_DISK_GB: f64 = 8.0;
282const GIB: f64 = 1024.0 * 1024.0 * 1024.0;
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look283/// The Durable Object behind each container is billed for as long as the
284/// container runs, at 128 MB.
285const SANDBOX_DO_GB: f64 = 0.125;
The keeper reads Cloudflare as it really answers286
287/// Cloudflare's published Containers rates, in dollars, used for any rate
288/// the bill does not show yet (while usage is inside the included amount).
289const LIST_MEMORY_GIB_SECOND: f64 = 0.000_002_5;
290const LIST_DISK_GB_SECOND: f64 = 0.000_000_07;
291const LIST_VCPU_SECOND: f64 = 0.000_02;
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look292/// Durable Objects duration: $12.50 per million GB-seconds.
293const LIST_DO_GB_SECOND: f64 = 0.000_012_5;
The keeper reads Cloudflare as it really answers294
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look295/// What one second of a sandbox costs whatever it does, in millionths of a
296/// dollar: its memory and disk, and the Durable Object behind it, for the
297/// whole second. CPU is billed only while busy, on top.
298pub(crate) fn sandbox_base_micros(memory: f64, disk: f64, durable_object: f64) -> f64 {
299 (SANDBOX_GIB * memory + SANDBOX_DISK_GB * disk + SANDBOX_DO_GB * durable_object) * MICROS_PER_DOLLAR as f64
300}
301
302/// What one second of a sandbox costs on average, in millionths of a
303/// dollar: its base, and the CPU sandboxes actually use per second of
304/// running. Runs that report their own CPU are priced on it instead (see
305/// `run_cost`).
306pub(crate) fn sandbox_second_micros(usage: ContainerUsage, memory: f64, disk: f64, vcpu: f64, durable_object: f64) -> Option<f64> {
The keeper reads Cloudflare as it really answers307 let instance_seconds = usage.memory_byte_seconds / (SANDBOX_GIB * GIB);
308 if instance_seconds < 3600.0 {
309 return None;
310 }
311 let cpu_share = usage.cpu_seconds / instance_seconds;
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look312 Some(sandbox_base_micros(memory, disk, durable_object) + cpu_share * vcpu * MICROS_PER_DOLLAR as f64)
313}
314
Fast pages, required checks on the branch, self-hosted runners, honest incidents315/// How much more a second of a larger machine's memory and disk (and the
316/// Durable Object behind it) costs than the standard sandbox's, at
317/// Cloudflare's list rates: 1 for the standard machine.
318pub(crate) fn base_scale(memory_gib: f64, disk_gb: f64) -> f64 {
319 let base = |memory: f64, disk: f64| memory * LIST_MEMORY_GIB_SECOND + disk * LIST_DISK_GB_SECOND + SANDBOX_DO_GB * LIST_DO_GB_SECOND;
320 base(memory_gib, disk_gb) / base(SANDBOX_GIB, SANDBOX_DISK_GB)
321}
322
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look323/// What a run that reported its own CPU cost g1t: its base for every
324/// second, and its vCPU-seconds at the vCPU rate.
325pub(crate) fn run_cost(seconds: i64, cpu_seconds: f64, base_per_second: f64, per_vcpu_second: f64) -> f64 {
326 seconds.max(0) as f64 * base_per_second + cpu_seconds.max(0.0) * per_vcpu_second
The keeper reads Cloudflare as it really answers327}
328
329/// A unit's marginal rate from the bill: the median, over the days that
330/// were charged, of cost over quantity. None while nothing was charged.
331pub(crate) fn billed_rate(rows: &[&UsageRow]) -> Option<f64> {
332 let mut rates: Vec<f64> = rows
333 .iter()
334 .filter(|r| r.cost > 0.0 && r.quantity > 0.0)
335 .map(|r| r.cost / r.quantity)
336 .collect();
337 if rates.is_empty() {
338 return None;
339 }
340 rates.sort_by(f64::total_cmp);
341 Some(rates[rates.len() / 2])
342}
343
Prices keep themselves current with what g1t pays344/// One line of Cloudflare's billable usage.
345#[derive(Debug, Clone)]
346pub(crate) struct UsageRow {
347 period_start: String,
348 period_end: String,
349 service: String,
350 unit: String,
351 quantity: f64,
352 cost: f64,
353}
354
355impl UsageRow {
356 /// Read leniently: the API is new, and its field names are FOCUS's.
357 fn from_value(row: &Value) -> Option<Self> {
358 let text = |keys: &[&str]| keys.iter().find_map(|k| row[*k].as_str()).unwrap_or_default().to_owned();
359 let number = |keys: &[&str]| {
360 keys.iter()
361 .find_map(|k| row[*k].as_f64().or_else(|| row[*k].as_str().and_then(|s| s.parse().ok())))
362 .unwrap_or(0.0)
363 };
364 let service = text(&["ServiceName", "service_name", "service"]);
365 if service.is_empty() {
366 return None;
367 }
368 let family = text(&["ServiceFamilyName", "service_family_name"]);
369 Some(UsageRow {
370 period_start: text(&["ChargePeriodStart", "charge_period_start"]),
371 period_end: text(&["ChargePeriodEnd", "charge_period_end"]),
372 service: if family.is_empty() { service } else { format!("{family} / {service}") },
The keeper reads Cloudflare as it really answers373 unit: text(&["PricingUnit", "ConsumedUnit", "consumed_unit"]),
Prices keep themselves current with what g1t pays374 quantity: number(&["PricingQuantity", "ConsumedQuantity", "pricing_quantity"]),
The keeper reads Cloudflare as it really answers375 // What g1t pays; list price if nothing was contracted.
376 cost: Some(number(&["ContractedCost", "BilledCost", "contracted_cost"]))
377 .filter(|cost| *cost > 0.0)
378 .unwrap_or_else(|| number(&["ListCost", "list_cost"])),
Prices keep themselves current with what g1t pays379 })
380 }
381}
382
383#[derive(Deserialize)]
384struct PriceRow {
385 meter: String,
386 title: String,
387 unit: String,
388 cost_micros: f64,
389 markup_percent: u32,
390 source: String,
391 checked_at: Option<String>,
392 updated_at: String,
393}
394
395#[derive(Deserialize)]
396struct ChangeRow {
397 meter: String,
398 old_cost_micros: f64,
399 new_cost_micros: f64,
400 markup_percent: u32,
Prices are what g1t pays plus 20%, from the first second401 old_markup_percent: Option<u32>,
Prices keep themselves current with what g1t pays402 reason: String,
403 created_at: String,
404}
405
406#[derive(Deserialize)]
407struct Unsettled {
408 id: String,
409 workspace: String,
410 repo: String,
411 number: u32,
412 task: String,
413 model: String,
414 token_hash: String,
415 billed_to: Option<String>,
416 session_id: String,
417 created_at: String,
418 finished_at: Option<String>,
419}
420
421#[derive(Deserialize)]
422struct Charged {
423 cost_micros: Option<i64>,
424 description: String,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put425 amount_micros: i64,
426}
427
428/// What a settled run's correction comes to, from the charges at the
429/// reported and the gateway's cost and what the workspace was charged at
430/// first. A charge up is drawn down like any charge; a charge down is
431/// given back only up to what the workspace paid, since what a credit or
432/// a pool paid was never the workspace's money.
433pub(crate) fn correction(reported_charge: i64, gateway_charge: i64, first_charged: i64) -> i64 {
434 let delta = gateway_charge - reported_charge;
435 if delta >= 0 { delta } else { delta.max(-first_charged.max(0)) }
Prices keep themselves current with what g1t pays436}
437
Merge branch 'worktree-agent-a633ac0f7f66d419d'438/// A session's logs are read 50 at a time, up to this many pages.
439const MAX_LOG_PAGES: u32 = 40;
440
441/// What AI Gateway's logs say a session cost.
442#[derive(Clone, Debug, Default, PartialEq)]
443pub(crate) struct SessionCost {
444 /// What the gateway priced the requests at, in dollars.
445 pub cost_usd: f64,
446 pub requests: u32,
447 /// Requests that used tokens but that the gateway put no price on: a
448 /// model it has no price for. Their cost is not in `cost_usd`.
449 pub unpriced: u32,
450 /// The models of those, for the statement and the drift.
451 pub unpriced_models: Vec<String>,
452 /// False when there were more logs than were read.
453 pub complete: bool,
454}
455
456impl SessionCost {
457 /// Adds one log, read leniently: `cost` in dollars, `tokens_in` and
458 /// `tokens_out`, `cached` for an answer from the gateway's own cache
459 /// (which costs nothing).
460 pub(crate) fn add(&mut self, log: &Value) {
461 self.requests += 1;
462 let number = |key: &str| log[key].as_f64().or_else(|| log[key].as_str().and_then(|s| s.parse().ok()));
463 let cost = number("cost").filter(|c| c.is_finite() && *c > 0.0);
464 let tokens = number("tokens_in").unwrap_or(0.0) + number("tokens_out").unwrap_or(0.0);
465 let cached = log["cached"].as_bool().unwrap_or(false);
466 match cost {
467 Some(cost) => self.cost_usd += cost,
468 None if tokens > 0.0 && !cached => {
469 self.unpriced += 1;
470 let model = log["model"].as_str().unwrap_or("an unnamed model").to_owned();
471 if !self.unpriced_models.contains(&model) {
472 self.unpriced_models.push(model);
473 }
474 }
475 None => {}
476 }
477 }
478
479 /// Whether the gateway's figure is the whole of what the run cost.
480 pub(crate) fn whole(&self) -> bool {
481 self.complete && self.unpriced == 0
482 }
483}
484
485/// The cost a run is settled at, in millionths: the gateway's figure when
486/// it priced every request; otherwise (a model it has no price for, or
487/// more logs than were read) never less than the sandbox reported, since
488/// the gateway's sum is then short of what the provider bills. With a
489/// reason for the statement and the drift when it is not the gateway's
490/// figure alone.
491pub(crate) fn settled_cost(reported_micros: i64, gateway: &SessionCost) -> (i64, Option<String>) {
492 // Not held to MAX_RUN_COST_USD: the gateway's figure is trusted.
493 let priced = if gateway.cost_usd.is_finite() { (gateway.cost_usd.max(0.0) * MICROS_PER_DOLLAR as f64).ceil() as i64 } else { 0 };
494 if gateway.whole() {
495 return (priced, None);
496 }
497 let mut why = Vec::new();
498 if gateway.unpriced > 0 {
499 why.push(format!(
500 "AI Gateway has no price for {} of its {} requests ({})",
501 gateway.unpriced,
502 gateway.requests,
503 gateway.unpriced_models.join(", ")
504 ));
505 }
506 if !gateway.complete {
507 why.push(format!("more than {} of its requests were logged", gateway.requests));
508 }
509 (priced.max(reported_micros.max(0)), Some(why.join("; ")))
510}
511
Prices keep themselves current with what g1t pays512fn ms(timestamp: &str) -> u64 {
513 // RFC 3339 in UTC, as g1t writes them.
514 worker::js_sys::Date::parse(timestamp) as u64
515}
516
517impl Billing {
518 pub(crate) async fn prices(&self) -> Result<PriceBook> {
Billing answers every page in a few round trips, not forty: its reads run together, Stripe is asked at most hourly, and the ledger has a time index519 // Three reads at once; the plans are priced from the rows already
520 // read, not one query per meter (status.g1t.sh times this call).
521 let (prices, changes, coming) = futures_util::future::join3(
522 async { self.db.prepare("SELECT * FROM prices ORDER BY rowid").all().await?.results::<PriceRow>() },
523 async {
524 self.db
525 .prepare("SELECT * FROM price_changes ORDER BY created_at DESC LIMIT 20")
526 .all()
527 .await?
528 .results::<ChangeRow>()
529 },
530 // Changes still to come first, so a rise is seen before it is charged.
531 self.coming_changes(),
532 )
533 .await;
534 let (prices, changes, coming) = (prices?, changes?, coming?);
Usage, Billing settings and prepaid AI credit; fixes from the UX audit535 let model_markup = prices.iter().find(|row| row.meter == "agent_models").map_or(self.margin_percent, |row| row.markup_percent);
Billing answers every page in a few round trips, not forty: its reads run together, Stripe is asked at most hourly, and the ledger has a time index536 let book: std::collections::BTreeMap<&str, f64> = prices
537 .iter()
538 .map(|row| (row.meter.as_str(), Price::price_for(row.cost_micros, row.markup_percent)))
539 .collect();
Merge Stripe Tax, the card fee on card payments, and one free workspace per person540 // With the card fee on top of each monthly price, as it is charged.
541 let card_fee = self.card_fee().await?;
Teams and CODEOWNERS, labels and milestones, dependency updates, the security suite, and a clearer top bar542 let plans: Vec<_> = g1t_contracts::billing::Feature::ALL
543 .iter()
Merge Stripe Tax, the card fee on card payments, and one free workspace per person544 .map(|feature| {
545 let mut plan = match feature {
546 g1t_contracts::billing::Feature::Security => crate::features::security_plan_at(&book),
547 _ => self.plan_at(&book),
548 };
549 plan.card_fee_cents = crate::tax::fee_for(i64::from(plan.monthly_cents), &card_fee) as u32;
550 plan
Teams and CODEOWNERS, labels and milestones, dependency updates, the security suite, and a clearer top bar551 })
552 .collect();
Prices keep themselves current with what g1t pays553 Ok(PriceBook {
554 prices: prices
555 .into_iter()
556 .map(|row| Price {
557 price_micros: Price::price_for(row.cost_micros, row.markup_percent),
558 meter: row.meter,
559 title: row.title,
560 unit: row.unit,
561 cost_micros: row.cost_micros,
562 markup_percent: row.markup_percent,
563 source: row.source,
564 checked_at: row.checked_at,
565 updated_at: row.updated_at,
566 })
567 .collect(),
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily568 changes: coming
Prices keep themselves current with what g1t pays569 .into_iter()
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily570 .chain(changes.into_iter().map(|row| PriceChange {
Prices keep themselves current with what g1t pays571 meter: row.meter,
572 old_cost_micros: row.old_cost_micros,
573 new_cost_micros: row.new_cost_micros,
574 markup_percent: row.markup_percent,
Prices are what g1t pays plus 20%, from the first second575 old_markup_percent: row.old_markup_percent,
Prices keep themselves current with what g1t pays576 reason: row.reason,
577 created_at: row.created_at,
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily578 effective_at: None,
579 }))
Prices keep themselves current with what g1t pays580 .collect(),
Usage, Billing settings and prepaid AI credit; fixes from the UX audit581 // The markup on models' provider price: the price book's
582 // `agent_models` (none from 2026-10-08).
583 model_margin_percent: model_markup,
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily584 plans,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put585 free: Some(g1t_contracts::billing::FreeTier {
586 trial_workspace_micros: if self.trials_on { self.plans.trial_workspace_micros } else { 0 },
587 trial_monthly_pool_micros: if self.trials_on { self.plans.trial_monthly_pool_micros } else { 0 },
588 oss_pool_micros: self.plans.oss_pool_micros,
589 oss_repo_micros: self.plans.oss_repo_micros,
590 free_private_storage_bytes: self.plans.free_storage_bytes,
Audit logs are kept by plan: a week on free, 90 days on the plan, and what staff set for an account in sudo591 audit_retention_days: self.plans.free_audit_days,
592 plan_audit_retention_days: self.plans.audit_days,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put593 min_charge_micros: self.plans.min_charge_micros,
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look594 git_operations_included: self.plans.git_included,
595 paid_start_ceiling_micros: self.plans.paid_start_micros,
596 overage_forgive_cost_micros: self.plans.forgive_cost_micros,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put597 }),
Prices keep themselves current with what g1t pays598 })
599 }
600
The keeper reads Cloudflare as it really answers601 /// Whether the costs have never been checked against Cloudflare's bill.
602 pub(crate) async fn never_checked(&self) -> Result<bool> {
603 Ok(self
604 .db
605 .prepare("SELECT meter FROM prices WHERE checked_at IS NOT NULL LIMIT 1")
606 .first::<Value>(None)
607 .await?
608 .is_none())
609 }
610
Prices keep themselves current with what g1t pays611 /// A meter's cost and price per unit, from the book.
612 pub(crate) async fn price(&self, meter: &str) -> Result<Option<(f64, f64)>> {
613 #[derive(Deserialize)]
614 struct Row {
615 cost_micros: f64,
616 markup_percent: u32,
617 }
618 Ok(self
619 .db
620 .prepare("SELECT cost_micros, markup_percent FROM prices WHERE meter = ?")
621 .bind(&[meter.into()])?
622 .first::<Row>(None)
623 .await?
624 .map(|row| (row.cost_micros, Price::price_for(row.cost_micros, row.markup_percent))))
625 }
626
627 /// Corrects finished runs to what AI Gateway priced them at, and
628 /// charges runs whose sandbox died before reporting.
629 pub(crate) async fn settle_runs(&self, keeper: &Keeper) -> Result<()> {
630 if keeper.token.is_none() || keeper.gateway.is_empty() {
631 return Ok(());
632 }
633 let now = now_ms();
634 let runs = self
635 .db
636 .prepare(
637 "SELECT id, workspace, repo, number, task, model, token_hash, billed_to, session_id, created_at, finished_at
638 FROM runs
Merge branch 'model-routing'639 WHERE session_id IS NOT NULL AND settled_at IS NULL AND COALESCE(billed_to, 'g1t') = 'g1t'
Prices keep themselves current with what g1t pays640 AND ((finished_at IS NOT NULL AND finished_at < ?1) OR created_at < ?2)
641 ORDER BY created_at LIMIT 10",
642 )
643 .bind(&[rfc3339(now - SETTLE_AFTER_MS).into(), rfc3339(now - ABANDONED_AFTER_MS).into()])?
644 .all()
645 .await?
646 .results::<Unsettled>()?;
647 for run in runs {
Merge branch 'worktree-agent-a633ac0f7f66d419d'648 let gateway = match keeper.session_cost(&run.session_id).await {
Prices keep themselves current with what g1t pays649 Ok(found) => found,
650 Err(error) => {
651 worker::console_error!("could not read gateway logs for {}: {error}", run.id);
652 continue;
653 }
654 };
655 let since = ms(run.finished_at.as_deref().unwrap_or(&run.created_at));
Merge branch 'worktree-agent-a633ac0f7f66d419d'656 if gateway.requests == 0 && now.saturating_sub(since) < GIVE_UP_AFTER_MS {
Prices keep themselves current with what g1t pays657 continue;
658 }
Merge branch 'worktree-agent-a633ac0f7f66d419d'659 self.settle(&run, &gateway).await?;
Prices keep themselves current with what g1t pays660 }
661 Ok(())
662 }
663
Merge branch 'model-routing'664 /// Closes runs on a workspace's own model provider: none is on g1t's
665 /// gateway, so nothing is corrected, but tokens the proxy counted after
666 /// the run reported, or for a sandbox that died before reporting, are
667 /// charged their agent rate now. Needs no gateway token.
668 pub(crate) async fn settle_own_runs(&self) -> Result<()> {
669 #[derive(Deserialize)]
670 struct Own {
671 id: String,
672 workspace: String,
673 repo: String,
674 number: u32,
675 task: String,
676 model: String,
677 token_hash: String,
678 billed_to: Option<String>,
679 }
680 let now = now_ms();
681 let runs = self
682 .db
683 .prepare(
684 "SELECT id, workspace, repo, number, task, model, token_hash, billed_to
685 FROM runs
686 WHERE billed_to = 'workspace' AND session_id IS NOT NULL AND settled_at IS NULL
687 AND ((finished_at IS NOT NULL AND finished_at < ?1) OR created_at < ?2)
688 ORDER BY created_at LIMIT 25",
689 )
690 .bind(&[rfc3339(now - SETTLE_AFTER_MS).into(), rfc3339(now - ABANDONED_AFTER_MS).into()])?
691 .all()
692 .await?
693 .results::<Own>()?;
694 for run in runs {
695 let settled_at = rfc3339(now_ms());
696 let claimed = self
697 .db
698 .prepare(
699 "UPDATE runs SET settled_at = ?1, finished_at = COALESCE(finished_at, ?1)
700 WHERE id = ?2 AND settled_at IS NULL RETURNING id",
701 )
702 .bind(&[settled_at.as_str().into(), run.id.as_str().into()])?
703 .first::<Value>(None)
704 .await?;
705 if claimed.is_none() {
706 continue;
707 }
708 let row = RunRow {
709 workspace: run.workspace,
710 repo: run.repo,
711 number: run.number,
712 task: run.task,
713 model: run.model,
714 token_hash: run.token_hash,
715 billed_to: run.billed_to,
716 };
717 self.charge_agent_rate(&run.id, &row, None).await?;
718 }
719 Ok(())
720 }
721
Merge branch 'worktree-agent-a633ac0f7f66d419d'722 async fn settle(&self, run: &Unsettled, gateway: &SessionCost) -> Result<()> {
723 let requests = gateway.requests;
Prices keep themselves current with what g1t pays724 let row = RunRow {
725 workspace: run.workspace.clone(),
726 repo: run.repo.clone(),
727 number: run.number,
728 task: run.task.clone(),
729 model: run.model.clone(),
730 token_hash: run.token_hash.clone(),
731 billed_to: run.billed_to.clone(),
732 };
733 let charged = self
734 .db
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put735 .prepare("SELECT cost_micros, description, amount_micros FROM ledger WHERE reference = ?")
Prices keep themselves current with what g1t pays736 .bind(&[run.id.as_str().into()])?
737 .first::<Charged>(None)
738 .await?;
Merge branch 'worktree-agent-a633ac0f7f66d419d'739 let reported = charged.as_ref().and_then(|c| c.cost_micros).unwrap_or(0);
740 // The gateway's figure; never under what the sandbox reported when
741 // the gateway could not price all of it (see `settled_cost`).
742 let (gateway_micros, short) = settled_cost(reported, gateway);
Billing accounts, terms and enterprises; g1t is no longer free743 let terms = self.terms_of(&run.workspace).await?;
Usage, Billing settings and prepaid AI credit; fixes from the UX audit744 // Models at the price book's markup on the provider's price
745 // (`agent_models`), as `finish_run` charges them.
746 let markup = self.model_markup().await?;
Merge branch 'worktree-agent-a633ac0f7f66d419d'747 // A cost's charge on the account's terms, and what a discount gave
Usage, Billing settings and prepaid AI credit; fixes from the UX audit748 // below cost plus the markup (counted as given, see `charged`).
Prices keep themselves current with what g1t pays749 let charge_for = |micros: i64| {
Usage, Billing settings and prepaid AI credit; fixes from the UX audit750 if self.free { (0, 0) } else { terms.discounted(crate::margin_on(micros, markup)) }
Prices keep themselves current with what g1t pays751 };
752 let settled_at = rfc3339(now_ms());
753 // Claim it, so two crons never settle it twice.
754 let claimed = self
755 .db
Merge branch 'worktree-agent-a633ac0f7f66d419d'756 .prepare(
757 "UPDATE runs SET settled_at = ?, gateway_cost_micros = ?, gateway_note = ?, finished_at = COALESCE(finished_at, ?)
758 WHERE id = ? AND settled_at IS NULL RETURNING id",
759 )
Prices keep themselves current with what g1t pays760 .bind(&[
761 settled_at.as_str().into(),
762 (gateway_micros as f64).into(),
Merge branch 'worktree-agent-a633ac0f7f66d419d'763 short.as_deref().map_or(worker::wasm_bindgen::JsValue::NULL, Into::into),
Prices keep themselves current with what g1t pays764 settled_at.as_str().into(),
765 run.id.as_str().into(),
766 ])?
767 .first::<Value>(None)
768 .await?;
769 if claimed.is_none() || requests == 0 {
770 return Ok(());
771 }
Usage, Billing settings and prepaid AI credit; fixes from the UX audit772 // Tokens counted after the run reported are charged their agent
773 // rate now (ai.rs).
Merge branch 'model-routing'774 self.charge_agent_rate(&run.id, &row, None).await?;
Merge branch 'worktree-agent-a633ac0f7f66d419d'775 if let Some(why) = &short {
776 worker::console_warn!("run {} settled at no less than reported: {why}", run.id);
777 }
778 let short_note = short.as_ref().map_or(String::new(), |why| format!(" ({why}; charged at no less than the sandbox reported)"));
Prices keep themselves current with what g1t pays779 let free_note = if self.free { " (free while g1t is being built out)" } else { "" };
780 match charged {
781 // Never reported: charged now, from the gateway's figure.
782 None => {
Merge branch 'worktree-agent-a633ac0f7f66d419d'783 let (charge, discount) = charge_for(gateway_micros);
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look784 let eligible = crate::credits::eligible_for(Some(g1t_contracts::billing::ComputeKind::Agent), None);
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put785 let drawn = self.draw(&run.workspace, charge, &settled_at[..7], &eligible).await?;
Prices keep themselves current with what g1t pays786 let description = format!(
Merge branch 'worktree-agent-a633ac0f7f66d419d'787 "Work on {}#{}, settled from AI Gateway after the sandbox stopped without reporting{short_note}{free_note}{}",
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put788 run.repo,
789 run.number,
790 drawn.note()
Prices keep themselves current with what g1t pays791 );
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put792 self.enter(&run.workspace, EntryKind::Usage, -(charge - drawn.total()), &description, &run.id, Some(&row), Some(gateway_micros), None, None)
Prices keep themselves current with what g1t pays793 .await?;
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put794 self.record_drawn(&run.id, &drawn).await?;
Merge branch 'worktree-agent-a633ac0f7f66d419d'795 self.record_discount(&run.id, discount).await?;
Spend caps: a monthly budget for comped workspaces and a daily breaker on what g1t pays796 self.count_spend(&run.workspace, gateway_micros, charge - drawn.total(), &drawn).await;
Prices keep themselves current with what g1t pays797 }
798 Some(charged) => {
799 let delta = gateway_micros - reported;
800 if delta == 0 {
801 return Ok(());
802 }
Merge branch 'worktree-agent-a633ac0f7f66d419d'803 let ((was, was_given), (now, now_given)) = (charge_for(reported), charge_for(gateway_micros));
804 let change = correction(was, now, -charged.amount_micros);
805 // What the discount gives moves with the charge.
806 let discount = now_given - was_given;
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put807 // A charge up is paid for like any other charge.
808 let drawn = if change > 0 {
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look809 let eligible = crate::credits::eligible_for(Some(g1t_contracts::billing::ComputeKind::Agent), None);
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put810 self.draw(&run.workspace, change, &settled_at[..7], &eligible).await?
811 } else {
812 crate::credits::Drawn::default()
813 };
Prices keep themselves current with what g1t pays814 let description = format!(
Merge branch 'worktree-agent-a633ac0f7f66d419d'815 "Correction to “{}”: AI Gateway priced its {requests} model requests at {}, not {}{short_note}{}",
Prices keep themselves current with what g1t pays816 charged.description,
817 crate::features::dollars(gateway_micros),
818 crate::features::dollars(reported),
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put819 drawn.note(),
Prices keep themselves current with what g1t pays820 );
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put821 let reference = format!("{}/settled", run.id);
Prices keep themselves current with what g1t pays822 self.enter(
823 &run.workspace,
824 EntryKind::Usage,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put825 -(change - drawn.total()),
Prices keep themselves current with what g1t pays826 &description,
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put827 &reference,
Prices keep themselves current with what g1t pays828 Some(&row),
829 Some(delta),
830 None,
831 None,
832 )
833 .await?;
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put834 self.record_drawn(&reference, &drawn).await?;
Merge branch 'worktree-agent-a633ac0f7f66d419d'835 self.record_discount(&reference, discount).await?;
Spend caps: a monthly budget for comped workspaces and a daily breaker on what g1t pays836 self.count_spend(&run.workspace, delta, change - drawn.total(), &drawn).await;
Prices keep themselves current with what g1t pays837 }
838 }
839 Ok(())
840 }
841
842 /// Checks each cost against what Cloudflare billed this month, and
843 /// moves the ones that changed.
844 pub(crate) async fn reconcile(&self, keeper: &Keeper) -> Result<()> {
845 if keeper.token.is_none() {
846 return Ok(());
847 }
848 let now = rfc3339(now_ms());
849 let today = &now[..10];
The keeper reads Cloudflare as it really answers850 let since = rfc3339(now_ms() - 30 * 24 * 60 * 60 * 1000);
851 let rows = keeper.billable_usage(&since[..10], today).await?;
Prices keep themselves current with what g1t pays852 for row in &rows {
853 self.db
854 .prepare(
855 "INSERT INTO cloudflare_usage (period_start, period_end, service, unit, quantity, cost_usd, fetched_at)
856 VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7)
857 ON CONFLICT (period_start, service, unit) DO UPDATE SET
858 period_end = ?2, quantity = ?5, cost_usd = ?6, fetched_at = ?7",
859 )
860 .bind(&[
861 row.period_start.as_str().into(),
862 row.period_end.as_str().into(),
863 row.service.as_str().into(),
864 row.unit.as_str().into(),
865 row.quantity.into(),
866 row.cost.into(),
867 now.as_str().into(),
868 ])?
869 .run()
870 .await?;
871 }
872
The keeper reads Cloudflare as it really answers873 let named = |words: &[&str]| -> Vec<&UsageRow> {
Prices keep themselves current with what g1t pays874 rows.iter()
The keeper reads Cloudflare as it really answers875 .filter(|r| {
876 let service = r.service.to_lowercase();
877 words.iter().all(|word| service.contains(word))
878 })
879 .collect()
Prices keep themselves current with what g1t pays880 };
881
The keeper reads Cloudflare as it really answers882 // Containers: each resource at what the bill shows it costs, or
883 // the published rate while the included amount still covers it,
884 // over how much CPU g1t's sandboxes really use per second.
885 let memory = billed_rate(&named(&["container memory"]));
886 let disk = billed_rate(&named(&["container disk"]));
887 let vcpu = billed_rate(&named(&["container vcpu"]));
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look888 let durable_object = billed_rate(&named(&["durable objects", "duration"]));
The keeper reads Cloudflare as it really answers889 let usage = keeper.container_usage(&since[..10], today).await?;
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look890 let rates = (
The keeper reads Cloudflare as it really answers891 memory.unwrap_or(LIST_MEMORY_GIB_SECOND),
892 disk.unwrap_or(LIST_DISK_GB_SECOND),
893 vcpu.unwrap_or(LIST_VCPU_SECOND),
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look894 durable_object.unwrap_or(LIST_DO_GB_SECOND),
895 );
896 // The parts, for runs that report their own CPU.
897 let parts_reason = "Cloudflare's Containers and Durable Objects rates, as billed or published";
898 self.measure("sandbox_base_second", sandbox_base_micros(rates.0, rates.1, rates.3), parts_reason).await?;
899 self.measure("sandbox_cpu_second", rates.2 * MICROS_PER_DOLLAR as f64, parts_reason).await?;
900 if let Some(per_second) = sandbox_second_micros(usage, rates.0, rates.1, rates.2, rates.3) {
The keeper reads Cloudflare as it really answers901 let instance_seconds = usage.memory_byte_seconds / (SANDBOX_GIB * GIB);
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look902 let billed = [("memory", memory), ("disk", disk), ("vCPU", vcpu), ("Durable Object duration", durable_object)]
The keeper reads Cloudflare as it really answers903 .iter()
904 .filter(|(_, rate)| rate.is_some())
905 .map(|(name, _)| *name)
906 .collect::<Vec<_>>();
907 let reason = format!(
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look908 "Sandboxes used {:.2} vCPU per second over {:.0} hours of Cloudflare Containers in the last 30 days, with the Durable Object behind each; {}",
The keeper reads Cloudflare as it really answers909 usage.cpu_seconds / instance_seconds,
910 instance_seconds / 3600.0,
911 if billed.is_empty() {
912 "rates are Cloudflare's published ones".to_owned()
913 } else {
914 format!("{} at what Cloudflare billed", billed.join(", "))
915 },
916 );
917 for meter in ["sandbox_second", "build_second"] {
918 self.measure(meter, per_second, &reason).await?;
Prices keep themselves current with what g1t pays919 }
920 }
The keeper reads Cloudflare as it really answers921 // Apps run as Workers: per million requests and CPU milliseconds,
922 // once the bill shows them charged.
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put923 // Security scans' CPU follows the same Workers CPU rate.
924 let app_meters: [(&str, &[&str], &str); 3] = [
The keeper reads Cloudflare as it really answers925 ("app_requests", &["workers", "requests"], "requests"),
926 ("app_cpu", &["workers cpu"], "CPU ms"),
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put927 ("scan_cpu", &["workers cpu"], "CPU ms"),
The keeper reads Cloudflare as it really answers928 ];
929 for (meter, words, unit) in app_meters {
930 if let Some(rate) = billed_rate(&named(words)) {
931 let reason = format!("Cloudflare billed Workers {unit} at ${:.2} per million", rate * 1e6);
932 self.measure(meter, rate * 1e6 * MICROS_PER_DOLLAR as f64, &reason).await?;
Prices keep themselves current with what g1t pays933 }
934 }
935 self.db
936 .prepare("UPDATE prices SET checked_at = ?")
937 .bind(&[now.as_str().into()])?
938 .run()
939 .await?;
940 Ok(())
941 }
942
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily943 /// Proposes moving a meter's cost to a measurement (see `pricing`):
944 /// applied on its own when small, after notice when a rise; left for
945 /// staff when large or suspect.
Prices keep themselves current with what g1t pays946 async fn measure(&self, meter: &str, measured: f64, reason: &str) -> Result<()> {
Git storage hardened, pages in tens of milliseconds, honest security alerts, and costs reconciled daily947 if let Some(outcome) = self.propose(meter, measured, reason, "keeper").await? {
948 worker::console_log!("{meter}: {outcome}");
Prices keep themselves current with what g1t pays949 }
950 Ok(())
951 }
952}
953
954#[cfg(test)]
955mod tests {
956 use super::*;
957
958 #[test]
Team plan, an open-source pool, monthly trials and honest metering; the sidebar for everyone; a workspace that stays put959 fn a_correction_never_gives_back_what_the_workspace_did_not_pay() {
960 // Up by 2 cents: charged in full (then drawn down like any charge).
961 assert_eq!(correction(100_000, 120_000, 100_000), 20_000);
962 // Down by 2 cents, all of it paid by the workspace: given back.
963 assert_eq!(correction(120_000, 100_000, 120_000), -20_000);
964 // Down, but the open-source pool paid all but a cent: a cent back.
965 assert_eq!(correction(120_000, 100_000, 10_000), -10_000);
966 // Paid entirely by a credit or pool: nothing back.
967 assert_eq!(correction(120_000, 100_000, 0), 0);
Prices keep themselves current with what g1t pays968 }
969
Merge branch 'worktree-agent-a633ac0f7f66d419d'970 fn logs(entries: &[Value]) -> SessionCost {
971 let mut total = SessionCost { complete: true, ..SessionCost::default() };
972 for log in entries {
973 total.add(log);
974 }
975 total
976 }
977
978 #[test]
979 fn a_run_is_settled_at_the_gateways_figure_when_it_priced_every_request() {
980 let gateway = logs(&[
981 json!({ "cost": 0.012, "tokens_in": 4000, "tokens_out": 300, "model": "claude-sonnet-5-5" }),
982 json!({ "cost": "0.003", "tokens_in": 900, "tokens_out": 40, "model": "claude-haiku-4-5" }),
983 // Served from the gateway's own cache: no cost, and none owed.
984 json!({ "cost": 0, "tokens_in": 900, "tokens_out": 40, "cached": true }),
985 // An error with no tokens costs nothing either.
986 json!({ "cost": null, "tokens_in": 0, "tokens_out": 0 }),
987 ]);
988 assert!(gateway.whole());
989 assert_eq!(gateway.requests, 4);
990 // Down from what the sandbox said, or up: the gateway's figure.
991 assert_eq!(settled_cost(20_000, &gateway), (15_000, None));
992 assert_eq!(settled_cost(9_000, &gateway), (15_000, None));
993 }
994
995 #[test]
996 fn a_model_the_gateway_cannot_price_is_never_settled_down_to_nothing() {
997 let gateway = logs(&[
998 json!({ "cost": 0.002, "tokens_in": 100, "tokens_out": 10, "model": "claude-haiku-4-5" }),
999 json!({ "cost": 0, "tokens_in": 50_000, "tokens_out": 2_000, "model": "claude-new-1" }),
1000 json!({ "tokens_in": 50_000, "tokens_out": 2_000, "model": "claude-new-1" }),
1001 ]);
1002 assert!(!gateway.whole());
1003 assert_eq!((gateway.unpriced, gateway.unpriced_models.clone()), (2, vec!["claude-new-1".to_owned()]));
1004 // The sandbox said $0.90: kept, not cut to the gateway's $0.002.
1005 let (cost, why) = settled_cost(900_000, &gateway);
1006 assert_eq!(cost, 900_000);
1007 assert!(why.unwrap().contains("no price for 2 of its 3 requests (claude-new-1)"));
1008 // A sandbox that reported less than the gateway priced: the gateway's.
1009 assert_eq!(settled_cost(1_000, &gateway).0, 2_000);
1010 }
1011
1012 #[test]
1013 fn more_logs_than_were_read_never_settle_a_run_down() {
1014 let mut gateway = logs(&[json!({ "cost": 1.0, "tokens_in": 1, "tokens_out": 1 })]);
1015 gateway.complete = false;
1016 let (cost, why) = settled_cost(3_000_000, &gateway);
1017 assert_eq!(cost, 3_000_000);
1018 assert!(why.unwrap().contains("more than 1 of its requests"));
1019 }
1020
1021 #[test]
1022 fn a_gateway_figure_over_the_report_cap_is_charged_in_full() {
1023 // A sandbox's report is believed up to $100; the gateway's is not capped.
1024 let gateway = logs(&[json!({ "cost": 140.0, "tokens_in": 1, "tokens_out": 1 })]);
1025 assert_eq!(settled_cost(100_000_000, &gateway).0, 140_000_000);
1026 assert_eq!(crate::charge_micros(140.0, 20), 120_000_000);
1027 assert_eq!(crate::margin_on(140_000_000, 20), 168_000_000);
1028 // Exactly cost plus the margin, rounded up, in whole micros (dollars
1029 // as floats can come out a micro high), never under it.
1030 for cost in [0_i64, 1, 7, 999, 123_457, 99_999_999] {
1031 let exact = (cost * 120 + 99) / 100;
1032 assert_eq!(crate::margin_on(cost, 20), exact, "{cost}");
1033 assert!(crate::margin_on(cost, 20) * 100 >= cost * 120, "{cost}");
1034 assert!(crate::charge_micros(cost as f64 / 1e6, 20) >= exact, "{cost}");
1035 }
1036 }
1037
Prices keep themselves current with what g1t pays1038 #[test]
The keeper reads Cloudflare as it really answers1039 fn a_sandbox_second_is_its_memory_and_disk_and_the_cpu_it_uses() {
1040 // An hour of sandboxes that kept a fifth of a vCPU busy.
1041 let usage = ContainerUsage { cpu_seconds: 720.0, memory_byte_seconds: 3600.0 * 4.0 * GIB };
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look1042 let micros =
1043 sandbox_second_micros(usage, LIST_MEMORY_GIB_SECOND, LIST_DISK_GB_SECOND, LIST_VCPU_SECOND, LIST_DO_GB_SECOND).unwrap();
1044 // 4 x 2.5 + 8 x 0.07 + 0.125 x 12.5 + 0.2 x 20 = 16.1225
1045 assert!((micros - 16.1225).abs() < 1e-9, "{micros}");
1046 // The Durable Object adds about 11% to the second it left out.
1047 assert!((sandbox_base_micros(LIST_MEMORY_GIB_SECOND, LIST_DISK_GB_SECOND, LIST_DO_GB_SECOND) - 12.1225).abs() < 1e-9);
The keeper reads Cloudflare as it really answers1048 // Too little use to say anything.
Invite-only launch: sign in with GitHub, repository access and lifecycle, many emails, a new look1049 assert!(sandbox_second_micros(ContainerUsage { cpu_seconds: 1.0, memory_byte_seconds: GIB }, 1.0, 1.0, 1.0, 1.0).is_none());
1050 }
1051
1052 #[test]
1053 fn a_run_that_reports_its_cpu_is_priced_on_it() {
1054 let base = sandbox_base_micros(LIST_MEMORY_GIB_SECOND, LIST_DISK_GB_SECOND, LIST_DO_GB_SECOND);
1055 let vcpu = LIST_VCPU_SECOND * MICROS_PER_DOLLAR as f64;
1056 // A 10-minute cargo build that kept its half vCPU busy throughout.
1057 let heavy = run_cost(600, 300.0, base, vcpu);
1058 assert!((heavy - (600.0 * 12.1225 + 300.0 * 20.0)).abs() < 1e-6);
1059 // The same ten minutes, mostly idle, costs less.
1060 let light = run_cost(600, 30.0, base, vcpu);
1061 assert!(light < heavy);
1062 // The average would have under-priced the heavy one.
1063 let average = 600.0 * (base + 0.195 * vcpu);
1064 assert!(average < heavy && average > light);
1065 assert_eq!(run_cost(0, -1.0, base, vcpu), 0.0);
Fast pages, required checks on the branch, self-hosted runners, honest incidents1066 // A larger machine's base: its memory and disk, not its CPU.
1067 assert!((base_scale(SANDBOX_GIB, SANDBOX_DISK_GB) - 1.0).abs() < 1e-12);
1068 assert!((base_scale(12.0, 20.0) - 2.72).abs() < 0.01, "{}", base_scale(12.0, 20.0));
1069 assert!((base_scale(8.0, 16.0) - 1.87).abs() < 0.01, "{}", base_scale(8.0, 16.0));
The keeper reads Cloudflare as it really answers1070 }
1071
1072 #[test]
1073 fn a_billed_rate_is_the_median_of_the_charged_days() {
1074 let row = |quantity: f64, cost: f64| UsageRow {
1075 period_start: String::new(),
1076 period_end: String::new(),
1077 service: "Containers / Container Memory".into(),
1078 unit: "Count".into(),
1079 quantity,
1080 cost,
1081 };
1082 let rows = [row(100.0, 0.0), row(100.0, 0.0002), row(100.0, 0.00025), row(100.0, 0.00025)];
1083 assert_eq!(billed_rate(&rows.iter().collect::<Vec<_>>()), Some(0.000_002_5));
1084 assert_eq!(billed_rate(&[&row(5.0, 0.0)]), None);
1085 }
1086
1087 #[test]
Prices keep themselves current with what g1t pays1088 fn usage_rows_are_read_by_their_focus_names() {
1089 let row = UsageRow::from_value(&json!({
1090 "ServiceFamilyName": "Containers",
1091 "ServiceName": "Memory",
The keeper reads Cloudflare as it really answers1092 "PricingUnit": "GiB-seconds",
Prices keep themselves current with what g1t pays1093 "PricingQuantity": "1200.5",
1094 "ContractedCost": 0.003,
1095 "ChargePeriodStart": "2026-10-01",
1096 }))
1097 .unwrap();
1098 assert_eq!(row.service, "Containers / Memory");
1099 assert_eq!(row.quantity, 1200.5);
1100 assert_eq!(row.cost, 0.003);
1101 assert!(UsageRow::from_value(&json!({ "nothing": 1 })).is_none());
1102 }
1103}

This file's history is long; its oldest lines are credited to the oldest commit read.