@deeprelay/sdk 0.2.0 → 0.2.1

package/README.md

@@ -1,4 +1,4 @@
-# @deeprelay/sdk@0.2.0
+# @deeprelay/sdk@0.2.1
 
 A TypeScript SDK client for the api.deeprelay.ai API.
 
@@ -156,7 +156,7 @@ and is automatically generated by the
 [OpenAPI Generator](https://openapi-generator.tech) project:
 
 - API version: `1.0.0`
-- Package version: `0.2.0`
+- Package version: `0.2.1`
 - Generator version: `7.24.0`
 - Build package: `org.openapitools.codegen.languages.TypeScriptFetchClientCodegen`
 

package/dist/apis/UsageApi.d.ts

@@ -18,6 +18,8 @@ export interface ListUsageRequest {
     end?: Date;
     cursor?: string;
     limit?: number;
+    modality?: ListUsageModalityEnum;
+    model?: string;
 }
 /**
  *
@@ -28,10 +30,12 @@ export declare class UsageApi extends runtime.BaseAPI {
      */
     listUsageRequestOpts(requestParameters: ListUsageRequest): Promise<runtime.RequestOpts>;
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsageRaw(requestParameters: ListUsageRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<UsagePage>>;
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsage(requestParameters?: ListUsageRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<UsagePage>;
@@ -54,3 +58,13 @@ export declare const ListUsageGroupByEnum: {
     readonly GpuType: "gpu_type";
 };
 export type ListUsageGroupByEnum = typeof ListUsageGroupByEnum[keyof typeof ListUsageGroupByEnum];
+/**
+ * @export
+ */
+export declare const ListUsageModalityEnum: {
+    readonly Chat: "chat";
+    readonly Image: "image";
+    readonly Video: "video";
+    readonly Embedding: "embedding";
+};
+export type ListUsageModalityEnum = typeof ListUsageModalityEnum[keyof typeof ListUsageModalityEnum];

package/dist/apis/UsageApi.js

@@ -22,7 +22,7 @@ var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, ge
     });
 };
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.ListUsageGroupByEnum = exports.ListUsageBucketEnum = exports.UsageApi = void 0;
+exports.ListUsageModalityEnum = exports.ListUsageGroupByEnum = exports.ListUsageBucketEnum = exports.UsageApi = void 0;
 const runtime = require("../runtime");
 const UsagePage_1 = require("../models/UsagePage");
 /**
@@ -53,6 +53,12 @@ class UsageApi extends runtime.BaseAPI {
             if (requestParameters['limit'] != null) {
                 queryParameters['limit'] = requestParameters['limit'];
             }
+            if (requestParameters['modality'] != null) {
+                queryParameters['modality'] = requestParameters['modality'];
+            }
+            if (requestParameters['model'] != null) {
+                queryParameters['model'] = requestParameters['model'];
+            }
             const headerParameters = {};
             if (this.configuration && this.configuration.accessToken) {
                 const token = this.configuration.accessToken;
@@ -71,6 +77,7 @@ class UsageApi extends runtime.BaseAPI {
         });
     }
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsageRaw(requestParameters, initOverrides) {
@@ -81,6 +88,7 @@ class UsageApi extends runtime.BaseAPI {
         });
     }
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsage() {
@@ -107,3 +115,12 @@ exports.ListUsageGroupByEnum = {
     InstanceId: 'instance_id',
     GpuType: 'gpu_type'
 };
+/**
+ * @export
+ */
+exports.ListUsageModalityEnum = {
+    Chat: 'chat',
+    Image: 'image',
+    Video: 'video',
+    Embedding: 'embedding'
+};

package/dist/esm/apis/UsageApi.d.ts

@@ -18,6 +18,8 @@ export interface ListUsageRequest {
     end?: Date;
     cursor?: string;
     limit?: number;
+    modality?: ListUsageModalityEnum;
+    model?: string;
 }
 /**
  *
@@ -28,10 +30,12 @@ export declare class UsageApi extends runtime.BaseAPI {
      */
     listUsageRequestOpts(requestParameters: ListUsageRequest): Promise<runtime.RequestOpts>;
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsageRaw(requestParameters: ListUsageRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<UsagePage>>;
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsage(requestParameters?: ListUsageRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<UsagePage>;
@@ -54,3 +58,13 @@ export declare const ListUsageGroupByEnum: {
     readonly GpuType: "gpu_type";
 };
 export type ListUsageGroupByEnum = typeof ListUsageGroupByEnum[keyof typeof ListUsageGroupByEnum];
+/**
+ * @export
+ */
+export declare const ListUsageModalityEnum: {
+    readonly Chat: "chat";
+    readonly Image: "image";
+    readonly Video: "video";
+    readonly Embedding: "embedding";
+};
+export type ListUsageModalityEnum = typeof ListUsageModalityEnum[keyof typeof ListUsageModalityEnum];

package/dist/esm/apis/UsageApi.js

@@ -50,6 +50,12 @@ export class UsageApi extends runtime.BaseAPI {
             if (requestParameters['limit'] != null) {
                 queryParameters['limit'] = requestParameters['limit'];
             }
+            if (requestParameters['modality'] != null) {
+                queryParameters['modality'] = requestParameters['modality'];
+            }
+            if (requestParameters['model'] != null) {
+                queryParameters['model'] = requestParameters['model'];
+            }
             const headerParameters = {};
             if (this.configuration && this.configuration.accessToken) {
                 const token = this.configuration.accessToken;
@@ -68,6 +74,7 @@ export class UsageApi extends runtime.BaseAPI {
         });
     }
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsageRaw(requestParameters, initOverrides) {
@@ -78,6 +85,7 @@ export class UsageApi extends runtime.BaseAPI {
         });
     }
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope.
      * Time-bucketed usage
      */
     listUsage() {
@@ -103,3 +111,12 @@ export const ListUsageGroupByEnum = {
     InstanceId: 'instance_id',
     GpuType: 'gpu_type'
 };
+/**
+ * @export
+ */
+export const ListUsageModalityEnum = {
+    Chat: 'chat',
+    Image: 'image',
+    Video: 'video',
+    Embedding: 'embedding'
+};

package/dist/esm/models/UsageBucket.d.ts

@@ -10,41 +10,74 @@
  * Do not edit the class manually.
  */
 /**
+ * One `/usage` row. Every row carries `bucket_start`, `gpu_seconds` and `cost_cents`. When the request set `modality` or `model` the row is an inference-usage row: it also carries `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count` (aggregating every serverless inference call in the bucket for that modality and model), and `gpu_seconds` is always 0. Otherwise it is an instance-usage row, carrying `instance_id` / `gpu_type` when `group_by` asked for them.
  *
  * @export
  * @interface UsageBucket
  */
 export interface UsageBucket {
     /**
-     *
+     * Start of the time bucket (truncated to `bucket`).
      * @type {Date}
      * @memberof UsageBucket
      */
     bucketStart: Date;
     /**
-     *
+     * Instance-usage rows with `group_by=instance_id` only.
      * @type {string}
      * @memberof UsageBucket
      */
     instanceId?: string;
     /**
-     *
+     * Instance-usage rows with `group_by=gpu_type` only.
      * @type {string}
      * @memberof UsageBucket
      */
     gpuType?: string;
     /**
-     *
+     * GPU instance seconds billed in the bucket. Always 0 on inference-usage rows.
      * @type {number}
      * @memberof UsageBucket
      */
     gpuSeconds: number;
     /**
-     *
+     * Total cost of the row, in US cents.
      * @type {number}
      * @memberof UsageBucket
      */
     costCents: number;
+    /**
+     * Inference-usage rows only: the modality of the calls in this row (chat, image, video or embedding).
+     *
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    modality?: string;
+    /**
+     * Inference-usage rows only: the model `id` the calls in this row were made against.
+     *
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    model?: string;
+    /**
+     * Inference-usage rows only. Sum of input tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    promptTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of output tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    completionTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of generated images; 0 for non-image modalities.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    imageCount?: number;
 }
 /**
  * Check if a given object implements the UsageBucket interface.

package/dist/esm/models/UsageBucket.js

@@ -36,6 +36,11 @@ export function UsageBucketFromJSONTyped(json, ignoreDiscriminator) {
         'gpuType': json['gpu_type'] == null ? undefined : json['gpu_type'],
         'gpuSeconds': json['gpu_seconds'],
         'costCents': json['cost_cents'],
+        'modality': json['modality'] == null ? undefined : json['modality'],
+        'model': json['model'] == null ? undefined : json['model'],
+        'promptTokens': json['prompt_tokens'] == null ? undefined : json['prompt_tokens'],
+        'completionTokens': json['completion_tokens'] == null ? undefined : json['completion_tokens'],
+        'imageCount': json['image_count'] == null ? undefined : json['image_count'],
     };
 }
 export function UsageBucketToJSON(json) {
@@ -51,5 +56,10 @@ export function UsageBucketToJSONTyped(value, ignoreDiscriminator = false) {
         'gpu_type': value['gpuType'],
         'gpu_seconds': value['gpuSeconds'],
         'cost_cents': value['costCents'],
+        'modality': value['modality'],
+        'model': value['model'],
+        'prompt_tokens': value['promptTokens'],
+        'completion_tokens': value['completionTokens'],
+        'image_count': value['imageCount'],
     };
 }

package/dist/esm/models/UsagePage.d.ts

@@ -17,6 +17,7 @@ import type { UsageBucket } from './UsageBucket';
  */
 export interface UsagePage {
     /**
+     * Inference-usage rows when the request set `modality` or `model`; instance-usage rows otherwise. One page never mixes the two.
      *
      * @type {Array<UsageBucket>}
      * @memberof UsagePage

package/dist/esm/runtime.js

@@ -61,7 +61,7 @@ export class Configuration {
         return undefined;
     }
     get headers() {
-        return Object.assign({ 'User-Agent': 'deeprelay-sdk-ts/0.2.0' }, this.configuration.headers);
+        return Object.assign({ 'User-Agent': 'deeprelay-sdk-ts/0.2.1' }, this.configuration.headers);
     }
     get credentials() {
         return this.configuration.credentials;

package/dist/models/UsageBucket.d.ts

@@ -10,41 +10,74 @@
  * Do not edit the class manually.
  */
 /**
+ * One `/usage` row. Every row carries `bucket_start`, `gpu_seconds` and `cost_cents`. When the request set `modality` or `model` the row is an inference-usage row: it also carries `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count` (aggregating every serverless inference call in the bucket for that modality and model), and `gpu_seconds` is always 0. Otherwise it is an instance-usage row, carrying `instance_id` / `gpu_type` when `group_by` asked for them.
  *
  * @export
  * @interface UsageBucket
  */
 export interface UsageBucket {
     /**
-     *
+     * Start of the time bucket (truncated to `bucket`).
      * @type {Date}
      * @memberof UsageBucket
      */
     bucketStart: Date;
     /**
-     *
+     * Instance-usage rows with `group_by=instance_id` only.
      * @type {string}
      * @memberof UsageBucket
      */
     instanceId?: string;
     /**
-     *
+     * Instance-usage rows with `group_by=gpu_type` only.
      * @type {string}
      * @memberof UsageBucket
      */
     gpuType?: string;
     /**
-     *
+     * GPU instance seconds billed in the bucket. Always 0 on inference-usage rows.
      * @type {number}
      * @memberof UsageBucket
      */
     gpuSeconds: number;
     /**
-     *
+     * Total cost of the row, in US cents.
      * @type {number}
      * @memberof UsageBucket
      */
     costCents: number;
+    /**
+     * Inference-usage rows only: the modality of the calls in this row (chat, image, video or embedding).
+     *
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    modality?: string;
+    /**
+     * Inference-usage rows only: the model `id` the calls in this row were made against.
+     *
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    model?: string;
+    /**
+     * Inference-usage rows only. Sum of input tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    promptTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of output tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    completionTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of generated images; 0 for non-image modalities.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    imageCount?: number;
 }
 /**
  * Check if a given object implements the UsageBucket interface.

package/dist/models/UsageBucket.js

@@ -43,6 +43,11 @@ function UsageBucketFromJSONTyped(json, ignoreDiscriminator) {
         'gpuType': json['gpu_type'] == null ? undefined : json['gpu_type'],
         'gpuSeconds': json['gpu_seconds'],
         'costCents': json['cost_cents'],
+        'modality': json['modality'] == null ? undefined : json['modality'],
+        'model': json['model'] == null ? undefined : json['model'],
+        'promptTokens': json['prompt_tokens'] == null ? undefined : json['prompt_tokens'],
+        'completionTokens': json['completion_tokens'] == null ? undefined : json['completion_tokens'],
+        'imageCount': json['image_count'] == null ? undefined : json['image_count'],
     };
 }
 function UsageBucketToJSON(json) {
@@ -58,5 +63,10 @@ function UsageBucketToJSONTyped(value, ignoreDiscriminator = false) {
         'gpu_type': value['gpuType'],
         'gpu_seconds': value['gpuSeconds'],
         'cost_cents': value['costCents'],
+        'modality': value['modality'],
+        'model': value['model'],
+        'prompt_tokens': value['promptTokens'],
+        'completion_tokens': value['completionTokens'],
+        'image_count': value['imageCount'],
     };
 }

package/dist/models/UsagePage.d.ts

@@ -17,6 +17,7 @@ import type { UsageBucket } from './UsageBucket';
  */
 export interface UsagePage {
     /**
+     * Inference-usage rows when the request set `modality` or `model`; instance-usage rows otherwise. One page never mixes the two.
      *
      * @type {Array<UsageBucket>}
      * @memberof UsagePage

package/dist/runtime.js

@@ -69,7 +69,7 @@ class Configuration {
         return undefined;
     }
     get headers() {
-        return Object.assign({ 'User-Agent': 'deeprelay-sdk-ts/0.2.0' }, this.configuration.headers);
+        return Object.assign({ 'User-Agent': 'deeprelay-sdk-ts/0.2.1' }, this.configuration.headers);
     }
     get credentials() {
         return this.configuration.credentials;

package/docs/UsageApi.md

@@ -10,10 +10,12 @@ All URIs are relative to *https://api.deeprelay.ai/v1*
 
 ## listUsage
 
-> UsagePage listUsage(bucket, groupBy, start, end, cursor, limit)
+> UsagePage listUsage(bucket, groupBy, start, end, cursor, limit, modality, model)
 
 Time-bucketed usage
 
+Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope. 
+
 ### Example
 
 ```ts
@@ -32,18 +34,22 @@ async function example() {
   const api = new UsageApi(config);
 
   const body = {
-    // 'hour' | 'day' | 'week' | 'month' (optional)
+    // 'hour' | 'day' | 'week' | 'month' | Bucket width. Defaults to `day`. (optional)
     bucket: bucket_example,
-    // 'instance_id' | 'gpu_type' (optional)
+    // 'instance_id' | 'gpu_type' | Split instance-usage rows by instance or GPU type. Applies only to the instance-usage shape (neither `modality` nor `model` set); ignored when either is set, since inference rows are always split by modality and model.  (optional)
     groupBy: groupBy_example,
-    // Date (optional)
+    // Date | Inclusive RFC 3339 range start. Defaults to 30 days ago. (optional)
     start: 2013-10-20T19:20:30+01:00,
-    // Date (optional)
+    // Date | Exclusive RFC 3339 range end. Defaults to now; must be after `start`. (optional)
     end: 2013-10-20T19:20:30+01:00,
     // string (optional)
     cursor: cursor_example,
     // number (optional)
     limit: 56,
+    // 'chat' | 'image' | 'video' | 'embedding' | Return inference-usage rows for this modality only. Setting `modality` or `model` selects the inference-usage shape; omit both for the instance-usage shape.  (optional)
+    modality: modality_example,
+    // string | Return inference-usage rows for this model only: an exact match on the model `id` as listed by `GET /models`. May be combined with `modality`. Setting `modality` or `model` selects the inference-usage shape.  (optional)
+    model: model_example,
   } satisfies ListUsageRequest;
 
   try {
@@ -63,12 +69,14 @@ example().catch(console.error);
 
 | Name | Type | Description  | Notes |
 |------------- | ------------- | ------------- | -------------|
-| **bucket** | `hour`, `day`, `week`, `month` |  | [Optional] [Defaults to `undefined`] [Enum: hour, day, week, month] |
-| **groupBy** | `instance_id`, `gpu_type` |  | [Optional] [Defaults to `undefined`] [Enum: instance_id, gpu_type] |
-| **start** | `Date` |  | [Optional] [Defaults to `undefined`] |
-| **end** | `Date` |  | [Optional] [Defaults to `undefined`] |
+| **bucket** | `hour`, `day`, `week`, `month` | Bucket width. Defaults to `day`. | [Optional] [Defaults to `undefined`] [Enum: hour, day, week, month] |
+| **groupBy** | `instance_id`, `gpu_type` | Split instance-usage rows by instance or GPU type. Applies only to the instance-usage shape (neither `modality` nor `model` set); ignored when either is set, since inference rows are always split by modality and model.  | [Optional] [Defaults to `undefined`] [Enum: instance_id, gpu_type] |
+| **start** | `Date` | Inclusive RFC 3339 range start. Defaults to 30 days ago. | [Optional] [Defaults to `undefined`] |
+| **end** | `Date` | Exclusive RFC 3339 range end. Defaults to now; must be after `start`. | [Optional] [Defaults to `undefined`] |
 | **cursor** | `string` |  | [Optional] [Defaults to `undefined`] |
 | **limit** | `number` |  | [Optional] [Defaults to `50`] |
+| **modality** | `chat`, `image`, `video`, `embedding` | Return inference-usage rows for this modality only. Setting `modality` or `model` selects the inference-usage shape; omit both for the instance-usage shape.  | [Optional] [Defaults to `undefined`] [Enum: chat, image, video, embedding] |
+| **model** | `string` | Return inference-usage rows for this model only: an exact match on the model `id` as listed by `GET /models`. May be combined with `modality`. Setting `modality` or `model` selects the inference-usage shape.  | [Optional] [Defaults to `undefined`] |
 
 ### Return type
 

package/docs/UsageBucket.md

@@ -1,6 +1,7 @@
 
 # UsageBucket
 
+One `/usage` row. Every row carries `bucket_start`, `gpu_seconds` and `cost_cents`. When the request set `modality` or `model` the row is an inference-usage row: it also carries `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count` (aggregating every serverless inference call in the bucket for that modality and model), and `gpu_seconds` is always 0. Otherwise it is an instance-usage row, carrying `instance_id` / `gpu_type` when `group_by` asked for them. 
 
 ## Properties
 
@@ -11,6 +12,11 @@ Name | Type
 `gpuType` | string
 `gpuSeconds` | number
 `costCents` | number
+`modality` | string
+`model` | string
+`promptTokens` | number
+`completionTokens` | number
+`imageCount` | number
 
 ## Example
 
@@ -24,6 +30,11 @@ const example = {
   "gpuType": null,
   "gpuSeconds": null,
   "costCents": null,
+  "modality": null,
+  "model": null,
+  "promptTokens": null,
+  "completionTokens": null,
+  "imageCount": null,
 } satisfies UsageBucket
 
 console.log(example)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@deeprelay/sdk",
-  "version": "0.2.0",
+  "version": "0.2.1",
   "description": "TypeScript SDK for the deeprelay Public Developer API",
   "author": "deeprelay Support <support@deeprelay.ai>",
   "repository": {

package/src/apis/UsageApi.ts

@@ -31,6 +31,8 @@ export interface ListUsageRequest {
     end?: Date;
     cursor?: string;
     limit?: number;
+    modality?: ListUsageModalityEnum;
+    model?: string;
 }
 
 /**
@@ -68,6 +70,14 @@ export class UsageApi extends runtime.BaseAPI {
             queryParameters['limit'] = requestParameters['limit'];
         }
 
+        if (requestParameters['modality'] != null) {
+            queryParameters['modality'] = requestParameters['modality'];
+        }
+
+        if (requestParameters['model'] != null) {
+            queryParameters['model'] = requestParameters['model'];
+        }
+
         const headerParameters: runtime.HTTPHeaders = {};
 
         if (this.configuration && this.configuration.accessToken) {
@@ -90,6 +100,7 @@ export class UsageApi extends runtime.BaseAPI {
     }
 
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope. 
      * Time-bucketed usage
      */
     async listUsageRaw(requestParameters: ListUsageRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<UsagePage>> {
@@ -100,6 +111,7 @@ export class UsageApi extends runtime.BaseAPI {
     }
 
     /**
+     * Returns spend aggregated into time buckets over `[start, end)` (default: the last 30 days, `bucket=day`), newest bucket first.  The response has one of two row shapes, selected by the query:  - **Inference usage** — set `modality` and/or `model`. Rows aggregate serverless inference calls, one row per (bucket, modality, model), and carry `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count`. This is the usage a serverless-inference customer is billed for. - **Instance usage** — neither `modality` nor `model` set. Rows aggregate GPU instance billing sessions, optionally split by `group_by`. With no instance usage this returns an empty `data` array, so to read inference spend always pass `modality` or `model`.  Requires the `billing:read` scope. 
      * Time-bucketed usage
      */
     async listUsage(requestParameters: ListUsageRequest = {}, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<UsagePage> {
@@ -127,3 +139,13 @@ export const ListUsageGroupByEnum = {
     GpuType: 'gpu_type'
 } as const;
 export type ListUsageGroupByEnum = typeof ListUsageGroupByEnum[keyof typeof ListUsageGroupByEnum];
+/**
+ * @export
+ */
+export const ListUsageModalityEnum = {
+    Chat: 'chat',
+    Image: 'image',
+    Video: 'video',
+    Embedding: 'embedding'
+} as const;
+export type ListUsageModalityEnum = typeof ListUsageModalityEnum[keyof typeof ListUsageModalityEnum];

package/src/models/UsageBucket.ts

@@ -14,41 +14,74 @@
 
 import { mapValues } from '../runtime';
 /**
+ * One `/usage` row. Every row carries `bucket_start`, `gpu_seconds` and `cost_cents`. When the request set `modality` or `model` the row is an inference-usage row: it also carries `modality`, `model`, `prompt_tokens`, `completion_tokens` and `image_count` (aggregating every serverless inference call in the bucket for that modality and model), and `gpu_seconds` is always 0. Otherwise it is an instance-usage row, carrying `instance_id` / `gpu_type` when `group_by` asked for them.
  * 
  * @export
  * @interface UsageBucket
  */
 export interface UsageBucket {
     /**
-     * 
+     * Start of the time bucket (truncated to `bucket`).
      * @type {Date}
      * @memberof UsageBucket
      */
     bucketStart: Date;
     /**
-     * 
+     * Instance-usage rows with `group_by=instance_id` only.
      * @type {string}
      * @memberof UsageBucket
      */
     instanceId?: string;
     /**
-     * 
+     * Instance-usage rows with `group_by=gpu_type` only.
      * @type {string}
      * @memberof UsageBucket
      */
     gpuType?: string;
     /**
-     * 
+     * GPU instance seconds billed in the bucket. Always 0 on inference-usage rows.
      * @type {number}
      * @memberof UsageBucket
      */
     gpuSeconds: number;
     /**
-     * 
+     * Total cost of the row, in US cents.
      * @type {number}
      * @memberof UsageBucket
      */
     costCents: number;
+    /**
+     * Inference-usage rows only: the modality of the calls in this row (chat, image, video or embedding).
+     * 
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    modality?: string;
+    /**
+     * Inference-usage rows only: the model `id` the calls in this row were made against.
+     * 
+     * @type {string}
+     * @memberof UsageBucket
+     */
+    model?: string;
+    /**
+     * Inference-usage rows only. Sum of input tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    promptTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of output tokens; 0 for modalities not billed per token.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    completionTokens?: number;
+    /**
+     * Inference-usage rows only. Sum of generated images; 0 for non-image modalities.
+     * @type {number}
+     * @memberof UsageBucket
+     */
+    imageCount?: number;
 }
 
 /**
@@ -76,6 +109,11 @@ export function UsageBucketFromJSONTyped(json: any, ignoreDiscriminator: boolean
         'gpuType': json['gpu_type'] == null ? undefined : json['gpu_type'],
         'gpuSeconds': json['gpu_seconds'],
         'costCents': json['cost_cents'],
+        'modality': json['modality'] == null ? undefined : json['modality'],
+        'model': json['model'] == null ? undefined : json['model'],
+        'promptTokens': json['prompt_tokens'] == null ? undefined : json['prompt_tokens'],
+        'completionTokens': json['completion_tokens'] == null ? undefined : json['completion_tokens'],
+        'imageCount': json['image_count'] == null ? undefined : json['image_count'],
     };
 }
 
@@ -95,6 +133,11 @@ export function UsageBucketToJSONTyped(value?: UsageBucket | null, ignoreDiscrim
         'gpu_type': value['gpuType'],
         'gpu_seconds': value['gpuSeconds'],
         'cost_cents': value['costCents'],
+        'modality': value['modality'],
+        'model': value['model'],
+        'prompt_tokens': value['promptTokens'],
+        'completion_tokens': value['completionTokens'],
+        'image_count': value['imageCount'],
     };
 }
 

package/src/models/UsagePage.ts

@@ -28,6 +28,7 @@ import {
  */
 export interface UsagePage {
     /**
+     * Inference-usage rows when the request set `modality` or `model`; instance-usage rows otherwise. One page never mixes the two.
      * 
      * @type {Array<UsageBucket>}
      * @memberof UsagePage

package/src/runtime.ts

@@ -75,7 +75,7 @@ export class Configuration {
     }
 
     get headers(): HTTPHeaders | undefined {
-        return { 'User-Agent': 'deeprelay-sdk-ts/0.2.0', ...this.configuration.headers };
+        return { 'User-Agent': 'deeprelay-sdk-ts/0.2.1', ...this.configuration.headers };
     }
 
     get credentials(): RequestCredentials | undefined {