@deeprelay/sdk 0.2.1 → 0.3.0

package/src/apis/InferenceApi.ts

@@ -2,7 +2,7 @@
 /* eslint-disable */
 /**
  * deeprelay Public Developer API
- * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details. 
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
  *
  * The version of the OpenAPI document: 1.0.0
  * Contact: support@deeprelay.ai
@@ -85,20 +85,24 @@ export interface CancelVideoRequest {
 
 export interface CreateChatCompletionRequest {
     chatCompletionRequest: ChatCompletionRequest;
+    xClientRequestId?: string;
 }
 
 export interface CreateEmbeddingsRequest {
     embeddingsRequest: EmbeddingsRequest;
+    xClientRequestId?: string;
 }
 
 export interface CreateImageRequest {
     imagesGenerationsRequest: ImagesGenerationsRequest;
     idempotencyKey?: string;
+    xClientRequestId?: string;
 }
 
 export interface CreateVideoRequest {
     videoCreateRequest: VideoCreateRequest;
     idempotencyKey?: string;
+    xClientRequestId?: string;
 }
 
 export interface GetModelRequest {
@@ -119,6 +123,7 @@ export interface InferencePreflightRequest {
 
 export interface ListModelsRequest {
     modality?: ListModelsModalityEnum;
+    ifNoneMatch?: string;
 }
 
 export interface ListVideosRequest {
@@ -201,6 +206,10 @@ export class InferenceApi extends runtime.BaseAPI {
 
         headerParameters['Content-Type'] = 'application/json';
 
+        if (requestParameters['xClientRequestId'] != null) {
+            headerParameters['X-Client-Request-Id'] = String(requestParameters['xClientRequestId']);
+        }
+
         if (this.configuration && this.configuration.accessToken) {
             const token = this.configuration.accessToken;
             const tokenString = await token("bearerAuth", []);
@@ -222,6 +231,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * OpenAI-compatible chat completion. Parameters fall into three groups:  - **Semantic parameters are gated per model.** `response_format`,   `tools`/`tool_choice`, `logprobs`/`top_logprobs` and `image_url`   content parts are sent only to a model whose `/v1/models` entry   lists the capability in `supported_parameters` (`vision` for   images). A model without it returns 400 `unsupported_parameter`   whose message names the models that do support it. When the model   supports the parameter but none of its currently routable upstreams   does, the request fails with 503 `upstream_unavailable` rather than   being silently degraded. - **Hints are forwarded or dropped, never rejected.**   `reasoning_effort`, `presence_penalty`, `frequency_penalty`, `seed`   and `parallel_tool_calls` are forwarded to an upstream that accepts   them and otherwise dropped. Every drop is reported in the   `x-deeprelay-ignored-params` response header. - **Rejected everywhere.** `logit_bias` and `function_call` return   400 `unsupported_parameter`; `input_audio` and `file` content parts   return 400 `unsupported_content_part`.  A non-strict `json_schema` sent to a model that can only guarantee JSON is served as `json_object` and reported as `response_format:downgraded`; a strict one is rejected. A `json_schema` over 64 KiB or nested deeper than 32 levels returns 400 `unsupported_parameter` (param `response_format`). An upstream rejection of a request the relay accepted returns 400 `upstream_rejected_request`. A non-chat model id returns 404 `model_not_found`. The request body is capped at 32 MiB (413 `request_too_large`).  Every successful response carries `usage.cost` and `usage.cost_details` (on a stream, only in the final usage chunk). See https://deeprelay.ai/docs/api/inference#openai-compatibility. 
      * Create a chat completion (OpenAI-compatible)
      */
     async createChatCompletionRaw(requestParameters: CreateChatCompletionRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<ChatCompletionResponse>> {
@@ -232,6 +242,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * OpenAI-compatible chat completion. Parameters fall into three groups:  - **Semantic parameters are gated per model.** `response_format`,   `tools`/`tool_choice`, `logprobs`/`top_logprobs` and `image_url`   content parts are sent only to a model whose `/v1/models` entry   lists the capability in `supported_parameters` (`vision` for   images). A model without it returns 400 `unsupported_parameter`   whose message names the models that do support it. When the model   supports the parameter but none of its currently routable upstreams   does, the request fails with 503 `upstream_unavailable` rather than   being silently degraded. - **Hints are forwarded or dropped, never rejected.**   `reasoning_effort`, `presence_penalty`, `frequency_penalty`, `seed`   and `parallel_tool_calls` are forwarded to an upstream that accepts   them and otherwise dropped. Every drop is reported in the   `x-deeprelay-ignored-params` response header. - **Rejected everywhere.** `logit_bias` and `function_call` return   400 `unsupported_parameter`; `input_audio` and `file` content parts   return 400 `unsupported_content_part`.  A non-strict `json_schema` sent to a model that can only guarantee JSON is served as `json_object` and reported as `response_format:downgraded`; a strict one is rejected. A `json_schema` over 64 KiB or nested deeper than 32 levels returns 400 `unsupported_parameter` (param `response_format`). An upstream rejection of a request the relay accepted returns 400 `upstream_rejected_request`. A non-chat model id returns 404 `model_not_found`. The request body is capped at 32 MiB (413 `request_too_large`).  Every successful response carries `usage.cost` and `usage.cost_details` (on a stream, only in the final usage chunk). See https://deeprelay.ai/docs/api/inference#openai-compatibility. 
      * Create a chat completion (OpenAI-compatible)
      */
     async createChatCompletion(requestParameters: CreateChatCompletionRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<ChatCompletionResponse> {
@@ -256,6 +267,10 @@ export class InferenceApi extends runtime.BaseAPI {
 
         headerParameters['Content-Type'] = 'application/json';
 
+        if (requestParameters['xClientRequestId'] != null) {
+            headerParameters['X-Client-Request-Id'] = String(requestParameters['xClientRequestId']);
+        }
+
         if (this.configuration && this.configuration.accessToken) {
             const token = this.configuration.accessToken;
             const tokenString = await token("bearerAuth", []);
@@ -317,6 +332,10 @@ export class InferenceApi extends runtime.BaseAPI {
             headerParameters['Idempotency-Key'] = String(requestParameters['idempotencyKey']);
         }
 
+        if (requestParameters['xClientRequestId'] != null) {
+            headerParameters['X-Client-Request-Id'] = String(requestParameters['xClientRequestId']);
+        }
+
         if (this.configuration && this.configuration.accessToken) {
             const token = this.configuration.accessToken;
             const tokenString = await token("bearerAuth", []);
@@ -378,6 +397,10 @@ export class InferenceApi extends runtime.BaseAPI {
             headerParameters['Idempotency-Key'] = String(requestParameters['idempotencyKey']);
         }
 
+        if (requestParameters['xClientRequestId'] != null) {
+            headerParameters['X-Client-Request-Id'] = String(requestParameters['xClientRequestId']);
+        }
+
         if (this.configuration && this.configuration.accessToken) {
             const token = this.configuration.accessToken;
             const tokenString = await token("bearerAuth", []);
@@ -454,6 +477,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * A key with a model allowlist gets 403 `model_not_allowed_for_key` for a model off its list, and the same 403 for an id that does not exist.
      * Get a specific model (OpenAI-compatible)
      */
     async getModelRaw(requestParameters: GetModelRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<Model>> {
@@ -464,6 +488,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * A key with a model allowlist gets 403 `model_not_allowed_for_key` for a model off its list, and the same 403 for an id that does not exist.
      * Get a specific model (OpenAI-compatible)
      */
     async getModel(requestParameters: GetModelRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<Model> {
@@ -649,6 +674,10 @@ export class InferenceApi extends runtime.BaseAPI {
 
         const headerParameters: runtime.HTTPHeaders = {};
 
+        if (requestParameters['ifNoneMatch'] != null) {
+            headerParameters['If-None-Match'] = String(requestParameters['ifNoneMatch']);
+        }
+
         if (this.configuration && this.configuration.accessToken) {
             const token = this.configuration.accessToken;
             const tokenString = await token("bearerAuth", []);
@@ -669,6 +698,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * Works with or without an API key, and returns the same body either way. Without a key the response is publicly cacheable (`Cache-Control: public, max-age=60`), carries a strong `ETag` (send `If-None-Match` to get a 304), and is rate-limited per client IP (429 `rate_limit_exceeded` with Retry-After). With a key it is `Cache-Control: private` and counts against the key\'s rate limit; an invalid key returns 401, it does not fall back to anonymous access. Error responses are `no-store`.  Each entry\'s `supported_parameters` is backed by live capability probes of the model\'s upstreams: it is the list to check before sending a gated parameter (`response_format`, `structured_outputs`, `logprobs`, `tools`, `vision`, ...).  One exception to \"the same body\": a key with a model allowlist (`allowed_models`) sees only the models on its list.  Each entry may carry a `metrics` object (always the last key): measured time to first token, decode rate and success rate over a rolling window of deeprelay customer traffic. Below the sample minimum `insufficient_data` is true and the figure is omitted, never zero. 
      * List available models (OpenAI-compatible)
      */
     async listModelsRaw(requestParameters: ListModelsRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<ModelList>> {
@@ -679,6 +709,7 @@ export class InferenceApi extends runtime.BaseAPI {
     }
 
     /**
+     * Works with or without an API key, and returns the same body either way. Without a key the response is publicly cacheable (`Cache-Control: public, max-age=60`), carries a strong `ETag` (send `If-None-Match` to get a 304), and is rate-limited per client IP (429 `rate_limit_exceeded` with Retry-After). With a key it is `Cache-Control: private` and counts against the key\'s rate limit; an invalid key returns 401, it does not fall back to anonymous access. Error responses are `no-store`.  Each entry\'s `supported_parameters` is backed by live capability probes of the model\'s upstreams: it is the list to check before sending a gated parameter (`response_format`, `structured_outputs`, `logprobs`, `tools`, `vision`, ...).  One exception to \"the same body\": a key with a model allowlist (`allowed_models`) sees only the models on its list.  Each entry may carry a `metrics` object (always the last key): measured time to first token, decode rate and success rate over a rolling window of deeprelay customer traffic. Below the sample minimum `insufficient_data` is true and the figure is omitted, never zero. 
      * List available models (OpenAI-compatible)
      */
     async listModels(requestParameters: ListModelsRequest = {}, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<ModelList> {

package/src/apis/MetaApi.ts

@@ -2,7 +2,7 @@
 /* eslint-disable */
 /**
  * deeprelay Public Developer API
- * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details. 
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
  *
  * The version of the OpenAPI document: 1.0.0
  * Contact: support@deeprelay.ai

package/src/apis/RequestsApi.ts

@@ -0,0 +1,335 @@
+/* tslint:disable */
+/* eslint-disable */
+/**
+ * deeprelay Public Developer API
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
+ *
+ * The version of the OpenAPI document: 1.0.0
+ * Contact: support@deeprelay.ai
+ *
+ * NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
+ * https://openapi-generator.tech
+ * Do not edit the class manually.
+ */
+
+import * as runtime from '../runtime';
+import {
+    type Problem,
+    ProblemFromJSON,
+    ProblemToJSON,
+} from '../models/Problem';
+import {
+    type RequestLogEntry,
+    RequestLogEntryFromJSON,
+    RequestLogEntryToJSON,
+} from '../models/RequestLogEntry';
+import {
+    type RequestLogPage,
+    RequestLogPageFromJSON,
+    RequestLogPageToJSON,
+} from '../models/RequestLogPage';
+
+export interface ExportRequestsRequest {
+    key?: string;
+    model?: string;
+    surface?: ExportRequestsSurfaceEnum;
+    status?: ExportRequestsStatusEnum;
+    errorCode?: string;
+    clientRequestId?: string;
+    from?: Date;
+    to?: Date;
+}
+
+export interface GetRequestRequest {
+    id: string;
+}
+
+export interface ListRequestsRequest {
+    key?: string;
+    model?: string;
+    surface?: ListRequestsSurfaceEnum;
+    status?: ListRequestsStatusEnum;
+    errorCode?: string;
+    clientRequestId?: string;
+    from?: Date;
+    to?: Date;
+    cursor?: string;
+    limit?: number;
+}
+
+/**
+ * 
+ */
+export class RequestsApi extends runtime.BaseAPI {
+
+    /**
+     * Creates request options for exportRequests without sending the request
+     */
+    async exportRequestsRequestOpts(requestParameters: ExportRequestsRequest): Promise<runtime.RequestOpts> {
+        const queryParameters: any = {};
+
+        if (requestParameters['key'] != null) {
+            queryParameters['key'] = requestParameters['key'];
+        }
+
+        if (requestParameters['model'] != null) {
+            queryParameters['model'] = requestParameters['model'];
+        }
+
+        if (requestParameters['surface'] != null) {
+            queryParameters['surface'] = requestParameters['surface'];
+        }
+
+        if (requestParameters['status'] != null) {
+            queryParameters['status'] = requestParameters['status'];
+        }
+
+        if (requestParameters['errorCode'] != null) {
+            queryParameters['error_code'] = requestParameters['errorCode'];
+        }
+
+        if (requestParameters['clientRequestId'] != null) {
+            queryParameters['client_request_id'] = requestParameters['clientRequestId'];
+        }
+
+        if (requestParameters['from'] != null) {
+            queryParameters['from'] = (requestParameters['from'] as any).toISOString();
+        }
+
+        if (requestParameters['to'] != null) {
+            queryParameters['to'] = (requestParameters['to'] as any).toISOString();
+        }
+
+        const headerParameters: runtime.HTTPHeaders = {};
+
+        if (this.configuration && this.configuration.accessToken) {
+            const token = this.configuration.accessToken;
+            const tokenString = await token("bearerAuth", []);
+
+            if (tokenString) {
+                headerParameters["Authorization"] = `Bearer ${tokenString}`;
+            }
+        }
+
+        let urlPath = `/requests/export`;
+
+        return {
+            path: urlPath,
+            method: 'GET',
+            headers: headerParameters,
+            query: queryParameters,
+        };
+    }
+
+    /**
+     * Streams every request-log row matching the filters as a CSV attachment, oldest columns first: `request_id, client_request_id, occurred_at, surface, model, key_id, key_name, key_prefix, status, http_status, error_code, finish_reason, prompt_tokens, completion_tokens, cached_tokens, reasoning_tokens, quota_units, cost, plan_covered, ttft_ms, duration_ms, ignored_params`. Takes the same filters and visibility rules as `GET /requests`, without the cursor and limit.  The export is counted before anything is written. When the filters match more rows than the export cap, the answer is 422 `export_too_large` and the problem body carries `max_rows`; narrow the time range or add a filter, then export again.  Requires the `serverless:read` scope. Exports have their own, tighter per-key rate limit. 
+     * Export the request log as CSV
+     */
+    async exportRequestsRaw(requestParameters: ExportRequestsRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<string>> {
+        const requestOptions = await this.exportRequestsRequestOpts(requestParameters);
+        const response = await this.request(requestOptions, initOverrides);
+
+        if (this.isJsonMime(response.headers.get('content-type'))) {
+            return new runtime.JSONApiResponse<string>(response);
+        } else {
+            return new runtime.TextApiResponse(response) as any;
+        }
+    }
+
+    /**
+     * Streams every request-log row matching the filters as a CSV attachment, oldest columns first: `request_id, client_request_id, occurred_at, surface, model, key_id, key_name, key_prefix, status, http_status, error_code, finish_reason, prompt_tokens, completion_tokens, cached_tokens, reasoning_tokens, quota_units, cost, plan_covered, ttft_ms, duration_ms, ignored_params`. Takes the same filters and visibility rules as `GET /requests`, without the cursor and limit.  The export is counted before anything is written. When the filters match more rows than the export cap, the answer is 422 `export_too_large` and the problem body carries `max_rows`; narrow the time range or add a filter, then export again.  Requires the `serverless:read` scope. Exports have their own, tighter per-key rate limit. 
+     * Export the request log as CSV
+     */
+    async exportRequests(requestParameters: ExportRequestsRequest = {}, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<string> {
+        const response = await this.exportRequestsRaw(requestParameters, initOverrides);
+        return await response.value();
+    }
+
+    /**
+     * Creates request options for getRequest without sending the request
+     */
+    async getRequestRequestOpts(requestParameters: GetRequestRequest): Promise<runtime.RequestOpts> {
+        if (requestParameters['id'] == null) {
+            throw new runtime.RequiredError(
+                'id',
+                'Required parameter "id" was null or undefined when calling getRequest().'
+            );
+        }
+
+        const queryParameters: any = {};
+
+        const headerParameters: runtime.HTTPHeaders = {};
+
+        if (this.configuration && this.configuration.accessToken) {
+            const token = this.configuration.accessToken;
+            const tokenString = await token("bearerAuth", []);
+
+            if (tokenString) {
+                headerParameters["Authorization"] = `Bearer ${tokenString}`;
+            }
+        }
+
+        let urlPath = `/requests/{id}`;
+        urlPath = urlPath.replace('{id}', encodeURIComponent(String(requestParameters['id'])));
+
+        return {
+            path: urlPath,
+            method: 'GET',
+            headers: headerParameters,
+            query: queryParameters,
+        };
+    }
+
+    /**
+     * Returns one request-log row. An unknown id, an id older than 90 days and an id the caller cannot see all answer the same 404. Requires the `serverless:read` scope. 
+     * Get one request
+     */
+    async getRequestRaw(requestParameters: GetRequestRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<RequestLogEntry>> {
+        const requestOptions = await this.getRequestRequestOpts(requestParameters);
+        const response = await this.request(requestOptions, initOverrides);
+
+        return new runtime.JSONApiResponse(response, (jsonValue) => RequestLogEntryFromJSON(jsonValue));
+    }
+
+    /**
+     * Returns one request-log row. An unknown id, an id older than 90 days and an id the caller cannot see all answer the same 404. Requires the `serverless:read` scope. 
+     * Get one request
+     */
+    async getRequest(requestParameters: GetRequestRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<RequestLogEntry> {
+        const response = await this.getRequestRaw(requestParameters, initOverrides);
+        return await response.value();
+    }
+
+    /**
+     * Creates request options for listRequests without sending the request
+     */
+    async listRequestsRequestOpts(requestParameters: ListRequestsRequest): Promise<runtime.RequestOpts> {
+        const queryParameters: any = {};
+
+        if (requestParameters['key'] != null) {
+            queryParameters['key'] = requestParameters['key'];
+        }
+
+        if (requestParameters['model'] != null) {
+            queryParameters['model'] = requestParameters['model'];
+        }
+
+        if (requestParameters['surface'] != null) {
+            queryParameters['surface'] = requestParameters['surface'];
+        }
+
+        if (requestParameters['status'] != null) {
+            queryParameters['status'] = requestParameters['status'];
+        }
+
+        if (requestParameters['errorCode'] != null) {
+            queryParameters['error_code'] = requestParameters['errorCode'];
+        }
+
+        if (requestParameters['clientRequestId'] != null) {
+            queryParameters['client_request_id'] = requestParameters['clientRequestId'];
+        }
+
+        if (requestParameters['from'] != null) {
+            queryParameters['from'] = (requestParameters['from'] as any).toISOString();
+        }
+
+        if (requestParameters['to'] != null) {
+            queryParameters['to'] = (requestParameters['to'] as any).toISOString();
+        }
+
+        if (requestParameters['cursor'] != null) {
+            queryParameters['cursor'] = requestParameters['cursor'];
+        }
+
+        if (requestParameters['limit'] != null) {
+            queryParameters['limit'] = requestParameters['limit'];
+        }
+
+        const headerParameters: runtime.HTTPHeaders = {};
+
+        if (this.configuration && this.configuration.accessToken) {
+            const token = this.configuration.accessToken;
+            const tokenString = await token("bearerAuth", []);
+
+            if (tokenString) {
+                headerParameters["Authorization"] = `Bearer ${tokenString}`;
+            }
+        }
+
+        let urlPath = `/requests`;
+
+        return {
+            path: urlPath,
+            method: 'GET',
+            headers: headerParameters,
+            query: queryParameters,
+        };
+    }
+
+    /**
+     * Returns one newest-first page of the organization\'s inference request log: one row per call to chat completions, messages, responses, embeddings, image generation or video creation, including calls that were refused before they reached a model (`status: rejected`). Rows are kept for 90 days; `window_days` says how far back the log reaches, and a `from` older than that is clamped rather than rejected.  Visibility follows the caller. An org admin\'s key and a `full_access` key see the whole organization; any other key sees only its own rows. A `key` filter naming a key the caller cannot see returns an empty page.  A row never carries prompt or completion content. `client_request_id` is the value the caller sent in `X-Client-Request-Id` (or `metadata.client_request_id` on `/responses`), kept for correlation.  Requires the `serverless:read` scope. Reads are rate-limited per key (429 `rate_limited` with Retry-After). 
+     * List requests (the request log)
+     */
+    async listRequestsRaw(requestParameters: ListRequestsRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<RequestLogPage>> {
+        const requestOptions = await this.listRequestsRequestOpts(requestParameters);
+        const response = await this.request(requestOptions, initOverrides);
+
+        return new runtime.JSONApiResponse(response, (jsonValue) => RequestLogPageFromJSON(jsonValue));
+    }
+
+    /**
+     * Returns one newest-first page of the organization\'s inference request log: one row per call to chat completions, messages, responses, embeddings, image generation or video creation, including calls that were refused before they reached a model (`status: rejected`). Rows are kept for 90 days; `window_days` says how far back the log reaches, and a `from` older than that is clamped rather than rejected.  Visibility follows the caller. An org admin\'s key and a `full_access` key see the whole organization; any other key sees only its own rows. A `key` filter naming a key the caller cannot see returns an empty page.  A row never carries prompt or completion content. `client_request_id` is the value the caller sent in `X-Client-Request-Id` (or `metadata.client_request_id` on `/responses`), kept for correlation.  Requires the `serverless:read` scope. Reads are rate-limited per key (429 `rate_limited` with Retry-After). 
+     * List requests (the request log)
+     */
+    async listRequests(requestParameters: ListRequestsRequest = {}, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<RequestLogPage> {
+        const response = await this.listRequestsRaw(requestParameters, initOverrides);
+        return await response.value();
+    }
+
+}
+
+/**
+ * @export
+ */
+export const ExportRequestsSurfaceEnum = {
+    Chat: 'chat',
+    Messages: 'messages',
+    Responses: 'responses',
+    Embeddings: 'embeddings',
+    Image: 'image',
+    Video: 'video'
+} as const;
+export type ExportRequestsSurfaceEnum = typeof ExportRequestsSurfaceEnum[keyof typeof ExportRequestsSurfaceEnum];
+/**
+ * @export
+ */
+export const ExportRequestsStatusEnum = {
+    Ok: 'ok',
+    Error: 'error',
+    Cancelled: 'cancelled',
+    Rejected: 'rejected'
+} as const;
+export type ExportRequestsStatusEnum = typeof ExportRequestsStatusEnum[keyof typeof ExportRequestsStatusEnum];
+/**
+ * @export
+ */
+export const ListRequestsSurfaceEnum = {
+    Chat: 'chat',
+    Messages: 'messages',
+    Responses: 'responses',
+    Embeddings: 'embeddings',
+    Image: 'image',
+    Video: 'video'
+} as const;
+export type ListRequestsSurfaceEnum = typeof ListRequestsSurfaceEnum[keyof typeof ListRequestsSurfaceEnum];
+/**
+ * @export
+ */
+export const ListRequestsStatusEnum = {
+    Ok: 'ok',
+    Error: 'error',
+    Cancelled: 'cancelled',
+    Rejected: 'rejected'
+} as const;
+export type ListRequestsStatusEnum = typeof ListRequestsStatusEnum[keyof typeof ListRequestsStatusEnum];

package/src/apis/UsageApi.ts

@@ -2,7 +2,7 @@
 /* eslint-disable */
 /**
  * deeprelay Public Developer API
- * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details. 
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
  *
  * The version of the OpenAPI document: 1.0.0
  * Contact: support@deeprelay.ai

package/src/apis/WebhooksApi.ts

@@ -2,7 +2,7 @@
 /* eslint-disable */
 /**
  * deeprelay Public Developer API
- * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details. 
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
  *
  * The version of the OpenAPI document: 1.0.0
  * Contact: support@deeprelay.ai

package/src/apis/index.ts

@@ -1,7 +1,10 @@
 /* tslint:disable */
 /* eslint-disable */
+export * from './AlertsApi';
+export * from './ApiKeysApi';
 export * from './BillingApi';
 export * from './InferenceApi';
 export * from './MetaApi';
+export * from './RequestsApi';
 export * from './UsageApi';
 export * from './WebhooksApi';

package/src/models/AlertEmailToggles.ts

@@ -0,0 +1,93 @@
+/* tslint:disable */
+/* eslint-disable */
+/**
+ * deeprelay Public Developer API
+ * Programmatic API for deeprelay. Authenticate with API keys minted from the dashboard. See https://deeprelay.ai/docs for details.  Keys minted from the dashboard or `deeprelay login` belong to the Production server, https://api.deeprelay.ai/v1. The Demo server is an internal staging environment with its own key store.  Versioning: `/v1` is the API version. `info.version` is the revision of this `/v1` document, not a separate API.  OpenAI compatibility: chat completions follow the OpenAI request and response shapes, but not every OpenAI parameter is carried. Some are rejected with 400 `unsupported_parameter` and any parameter not declared in this spec is accepted and ignored. The full table is at https://deeprelay.ai/docs/api/inference#openai-compatibility. 
+ *
+ * The version of the OpenAPI document: 1.0.0
+ * Contact: support@deeprelay.ai
+ *
+ * NOTE: This class is auto generated by OpenAPI Generator (https://openapi-generator.tech).
+ * https://openapi-generator.tech
+ * Do not edit the class manually.
+ */
+
+import { mapValues } from '../runtime';
+/**
+ * 
+ * @export
+ * @interface AlertEmailToggles
+ */
+export interface AlertEmailToggles {
+    /**
+     * 
+     * @type {boolean}
+     * @memberof AlertEmailToggles
+     */
+    balanceLow: boolean;
+    /**
+     * 
+     * @type {boolean}
+     * @memberof AlertEmailToggles
+     */
+    subscriptionQuota: boolean;
+    /**
+     * 
+     * @type {boolean}
+     * @memberof AlertEmailToggles
+     */
+    apiKeyBudget: boolean;
+    /**
+     * 
+     * @type {boolean}
+     * @memberof AlertEmailToggles
+     */
+    memberSpend: boolean;
+}
+
+/**
+ * Check if a given object implements the AlertEmailToggles interface.
+ */
+export function instanceOfAlertEmailToggles(value: object): value is AlertEmailToggles {
+    if ((!('balanceLow' in (value as Record<string, any>)) && !('balance_low' in (value as Record<string, any>))) || ((value as Record<string, any>)['balanceLow'] === undefined && (value as Record<string, any>)['balance_low'] === undefined)) return false;
+    if ((!('subscriptionQuota' in (value as Record<string, any>)) && !('subscription_quota' in (value as Record<string, any>))) || ((value as Record<string, any>)['subscriptionQuota'] === undefined && (value as Record<string, any>)['subscription_quota'] === undefined)) return false;
+    if ((!('apiKeyBudget' in (value as Record<string, any>)) && !('api_key_budget' in (value as Record<string, any>))) || ((value as Record<string, any>)['apiKeyBudget'] === undefined && (value as Record<string, any>)['api_key_budget'] === undefined)) return false;
+    if ((!('memberSpend' in (value as Record<string, any>)) && !('member_spend' in (value as Record<string, any>))) || ((value as Record<string, any>)['memberSpend'] === undefined && (value as Record<string, any>)['member_spend'] === undefined)) return false;
+    return true;
+}
+
+export function AlertEmailTogglesFromJSON(json: any): AlertEmailToggles {
+    return AlertEmailTogglesFromJSONTyped(json, false);
+}
+
+export function AlertEmailTogglesFromJSONTyped(json: any, ignoreDiscriminator: boolean): AlertEmailToggles {
+    if (json == null) {
+        return json;
+    }
+    return {
+        
+        'balanceLow': json['balance_low'],
+        'subscriptionQuota': json['subscription_quota'],
+        'apiKeyBudget': json['api_key_budget'],
+        'memberSpend': json['member_spend'],
+    };
+}
+
+export function AlertEmailTogglesToJSON(json: any): AlertEmailToggles {
+    return AlertEmailTogglesToJSONTyped(json, false);
+}
+
+export function AlertEmailTogglesToJSONTyped(value?: AlertEmailToggles | null, ignoreDiscriminator: boolean = false): any {
+    if (value == null) {
+        return value;
+    }
+
+    return {
+        
+        'balance_low': value['balanceLow'],
+        'subscription_quota': value['subscriptionQuota'],
+        'api_key_budget': value['apiKeyBudget'],
+        'member_spend': value['memberSpend'],
+    };
+}
+