@leaflow/sdk 0.0.0-dev.389.g87f5860 → 0.0.0-dev.391.g0aed14e

package/dist/billing/v1/generated/validators/withdrawalPreviewItem.zod.d.ts

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/billing/v1/generated/validators/withdrawalPreviewItem.zod.js

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/billing/v1/generated/validators/withdrawalPreviewRequest.zod.d.ts

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/billing/v1/generated/validators/withdrawalPreviewRequest.zod.js

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/billing/v1/generated/validators/withdrawalSource.zod.d.ts

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/billing/v1/generated/validators/withdrawalSource.zod.js

@@ -2,7 +2,7 @@
  * Generated by orval 🍺
  * Do not edit manually.
  * Leaflow Billing API
- * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, and, with project permissions granted in IAM, change its subscriptions (`billing:subscription.write`), change its billing account (`billing:project_assignment.write`) and read its costs (`billing:cost.read`). A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
+ * Products, prices, orders, subscriptions and account billing. Operations on a billing account require an account access token. The account, its balance, payment options and quotes are also readable by members of the projects it currently pays for; changing the account and moving its funds need the owner. Members of a project, as each operation states, can also read the project's billing account, subscriptions and orders, complete and pay for its purchases, quote purchases, changes, renewals and cancellations, and, with the project permission `billing:finance.manage` granted in IAM, renew, cancel and turn automatic renewal of its subscriptions on or off, change its billing account and read its costs. A member without the permission an operation needs is refused with 403 `PERMISSION_DENIED` and `meta.permission`; a caller who is not a member, with 403 `IAM_NOT_PROJECT_MEMBER`. Public catalogue operations declare no security requirement.
  * OpenAPI spec version: v1
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/functions.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import type { APIKeyResource, APIKeyUsageListResponseBody, ChatCompletion, ChatCompletionRequest, ConflictResponse, CreateAPIKeyRequestBody, CreateServiceRequestBody, CreateServiceResponseBody, CursorPageRequestResource, EmbeddingRequest, EmbeddingResponse, Error, GetUsageSummaryParams, GetUsageTimelineParams, InferenceError, IssuedAPIKeyResource, LengthAwarePageAPIKeyResource, ListApiKeysParams, ListRequestsParams, ListUsageByApiKeyParams, ListUsageByModelParams, Model, ModelList, ModelListResponseBody, ModelResource, ModelUsageListResponseBody, NativeRequest, NativeResponse, RequestResource, ResponseCreate, ResponseDeleted, ResponseResource, ServiceResource, StreamGenerateContentParams, TotalsResource, UpdateAPIKeyRequestBody, UsageTimelineResponseBody } from "./models.js";
@@ -439,7 +439,7 @@ export declare const getListModelsUrl: () => string;
  *
  * `context_length` and `max_output_tokens` are advisory. The server does not truncate on their basis, and the request body is forwarded upstream unchanged.
  *
- * `pricing` states what requests to each model cost in the currency of the project's current billing account.
+ * Canopy returns no amounts: each model's `billing` names the Billing plan and the rate dimensions its token usage is priced at. Read the rates from Billing with `search-prices` and `search-rates`.
  * @summary List the available models
  */
 export declare const listModels: (options?: RequestInit, fetchFn?: typeof globalThis.fetch) => Promise<listModelsResponse>;
@@ -528,7 +528,7 @@ export type getServiceResponseError = getServiceResponseDefault & {
 export type getServiceResponse = getServiceResponseSuccess | getServiceResponseError;
 export declare const getGetServiceUrl: () => string;
 /**
- * The model platform service of the authenticated project. The forwarding endpoints accept its requests only while `status` is `active`; each request is then billed under `subscription_id`, per token, at the rates stated in each model's `pricing`.
+ * The model platform service of the authenticated project. The forwarding endpoints accept its requests only while `status` is `active`; each request is then billed under `subscription_id`, per token, at the rates Billing holds for each model's `billing`.
  *
  * When the project is deleted, its API keys are revoked and the subscription is canceled; the service then reads `inactive` with `ended_at` set.
  * @summary Get the model platform service

package/dist/canopy/v1/generated/functions.js

@@ -440,7 +440,7 @@ export const getListModelsUrl = () => {
  *
  * `context_length` and `max_output_tokens` are advisory. The server does not truncate on their basis, and the request body is forwarded upstream unchanged.
  *
- * `pricing` states what requests to each model cost in the currency of the project's current billing account.
+ * Canopy returns no amounts: each model's `billing` names the Billing plan and the rate dimensions its token usage is priced at. Read the rates from Billing with `search-prices` and `search-rates`.
  * @summary List the available models
  */
 export const listModels = async (options, fetchFn) => {
@@ -518,7 +518,7 @@ export const getGetServiceUrl = () => {
     return `https://canopy.leaflow.cloud/api/v1/service`;
 };
 /**
- * The model platform service of the authenticated project. The forwarding endpoints accept its requests only while `status` is `active`; each request is then billed under `subscription_id`, per token, at the rates stated in each model's `pricing`.
+ * The model platform service of the authenticated project. The forwarding endpoints accept its requests only while `status` is `active`; each request is then billed under `subscription_id`, per token, at the rates Billing holds for each model's `billing`.
  *
  * When the project is deleted, its API keys are revoked and the subscription is canceled; the service then reads `inactive` with `ended_at` set.
  * @summary Get the model platform service

package/dist/canopy/v1/generated/models.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 export type ReasoningDiagnosticsRequested = {
@@ -325,8 +325,8 @@ export declare const ModelResourceStatus: {
     readonly draining: "draining";
     readonly retired: "retired";
 };
-export type TokenPriceTokenKind = (typeof TokenPriceTokenKind)[keyof typeof TokenPriceTokenKind];
-export declare const TokenPriceTokenKind: {
+export type ModelMeterTokenKind = (typeof ModelMeterTokenKind)[keyof typeof ModelMeterTokenKind];
+export declare const ModelMeterTokenKind: {
     readonly input: "input";
     readonly output: "output";
     readonly cache_read: "cache_read";
@@ -334,36 +334,30 @@ export declare const TokenPriceTokenKind: {
     readonly reasoning: "reasoning";
 };
 /**
- * A decimal string, in the currency stated alongside it.
- *
- * **The currency is not part of this type.** It is carried by a `currency` field next to the
- * amount, or by the account the amount belongs to. Reading an amount without that field is
- * reading a number with no unit.
- *
- * It is a string rather than a JSON number because a JSON number is a float in most parsers,
- * and a float loses precision on the first arithmetic. Nothing on this platform puts an amount
- * through a float.
+ * The dimension values the usage of this token kind carries, and the rate is matched at.
  */
-export type Money = string;
-export interface TokenPrice {
-    token_kind: TokenPriceTokenKind;
-    /** The charge for `unit_quantity` tokens of this kind. Null when the kind has no rate in the currency, in which case the model cannot be billed for it. */
-    unit_amount: Money | null;
-    /**
-     * How many tokens `unit_amount` is for, such as `1000000`. Null when `unit_amount` is null.
-     * @nullable
-     */
-    unit_quantity: string | null;
+export type ModelMeterDimensions = {
+    [key: string]: string;
+};
+export interface ModelMeter {
+    token_kind: ModelMeterTokenKind;
+    meter_lookup_key: string;
+    /** The dimension values the usage of this token kind carries, and the rate is matched at. */
+    dimensions: ModelMeterDimensions;
 }
 /**
- * What requests to a model cost, as Billing prices them in the currency of the project's current billing account. These are list amounts before promotion codes, not a checkout guarantee; usage is billed afterwards under the service's subscription.
+ * How requests to a model are billed: under the service's subscription, on the Billing plan `plan_id`, one rate for each token kind. Read the rates from Billing with `search-prices` for the plan and then `search-rates` with each meter's `dimensions`.
  *
  * Each token kind is priced on its own. `input` excludes cached input, which is `cache_read`, and `output` excludes reasoning the provider reports separately, which is `reasoning`; a request is billed for each kind it used.
  */
-export interface ModelPricing {
-    currency: string;
+export interface ModelBilling {
+    /**
+     * The Billing plan of the service; null where the service is not sold.
+     * @nullable
+     */
+    plan_id: string | null;
     /** One entry for each of the five token kinds, in the order input, output, cache_read, cache_write, reasoning. */
-    token_kinds: TokenPrice[];
+    meters: ModelMeter[];
 }
 export interface ModelResource {
     /** The context window. Advisory only; the server does not truncate on its basis */
@@ -375,8 +369,7 @@ export interface ModelResource {
     id: string;
     /** The maximum output length. Advisory only */
     max_output_tokens: number;
-    /** What requests to this model cost. Null when the project has no billing account. */
-    pricing: ModelPricing | null;
+    billing: ModelBilling;
     status: ModelResourceStatus;
     /** The model developer; not the upstream provider or model family. */
     developer: string;
@@ -550,7 +543,7 @@ export interface ServiceResource {
      */
     order_id: string | null;
     /**
-     * The Billing plan the service is sold under; null where the service is not sold. Price enabling it with Billing's quote, giving this plan with the `recurring` price type, the `metered` usage type and a quantity of 1; token usage is priced per model, as each model's `pricing` states.
+     * The Billing plan the service is sold under; null where the service is not sold. Price enabling it with Billing's quote, giving this plan with the `recurring` price type, the `metered` usage type and a quantity of 1; token usage is priced per model, at the rates Billing holds for each model's `billing`.
      * @nullable
      */
     plan_id: string | null;
@@ -584,6 +577,18 @@ export declare const CheckoutMode: {
     readonly automatic: "automatic";
     readonly deferred: "deferred";
 };
+/**
+ * A decimal string, in the currency stated alongside it.
+ *
+ * **The currency is not part of this type.** It is carried by a `currency` field next to the
+ * amount, or by the account the amount belongs to. Reading an amount without that field is
+ * reading a number with no unit.
+ *
+ * It is a string rather than a JSON number because a JSON number is a float in most parsers,
+ * and a float loses precision on the first arithmetic. Nothing on this platform puts an amount
+ * through a float.
+ */
+export type Money = string;
 /**
  * Shared checkout choices for a product purchase. The service may set them or forward its client's
  * choices unchanged; it does not need to read them, and Billing validates them. Omitting this object or

package/dist/canopy/v1/generated/models.js

@@ -38,7 +38,7 @@ export const ModelResourceStatus = {
     draining: "draining",
     retired: "retired",
 };
-export const TokenPriceTokenKind = {
+export const ModelMeterTokenKind = {
     input: "input",
     output: "output",
     cache_read: "cache_read",

package/dist/canopy/v1/generated/schemas.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/schemas.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyResource.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyResource.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyUsageListResponseBody.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyUsageListResponseBody.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyUsageResource.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/aPIKeyUsageResource.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/chatCompletion.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/chatCompletion.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/chatCompletionRequest.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/chatCompletionRequest.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/checkoutMode.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/checkoutMode.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/checkoutOptions.zod.d.ts

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";

package/dist/canopy/v1/generated/validators/checkoutOptions.zod.js

@@ -16,7 +16,7 @@
  * the model name and the credentials. Reasoning settings such as `reasoning_effort` are passed
  * through as written; a model's `capabilities` lists the values its upstream accepts.
  *
- * Management operations use `/api/v1`. The management model catalogue includes pricing;
+ * Management operations use `/api/v1`. The management model catalogue names how each model is billed in Billing;
  * `/v1/models` returns the OpenAI model format with Canopy capabilities and parameter limits.
  *
  * ## Authentication
@@ -35,7 +35,7 @@
  * - **A usage query must carry a time range**, spanning no more than 31 days, or 400 days for the usage timeline. `from` and `to` carry a timezone offset, which the server does not infer.
  * - **The request and response bodies do not appear in the ledger.** They are not recorded. Quote the `upstream_request_id` from the response when reporting a problem.
  * - **Usage may be estimated.** The upstream provider does not report usage on every request. Where it does not, `usage_source` is `estimated` and the figures are derived from the character classes of the payload, to within roughly fifteen per cent.
- * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates stated in each model's `pricing`.
+ * - **The service is enabled before use.** A project enables the model platform once, through `create-service`; the forwarding endpoints refuse its requests until the service is `active`. Requests are then billed under the service's subscription by the tokens they use, at the rates Billing holds for each model's `billing`.
  * OpenAPI spec version: 1.0.0
  */
 import * as zod from "zod";