@gpuai/sdk 0.5.1 → 0.5.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (36) hide show
  1. package/.openapi-generator/FILES +4 -0
  2. package/README.md +6 -2
  3. package/dist/apis/InferenceApi.d.ts +60 -0
  4. package/dist/apis/InferenceApi.js +126 -1
  5. package/dist/esm/apis/InferenceApi.d.ts +60 -0
  6. package/dist/esm/apis/InferenceApi.js +125 -0
  7. package/dist/esm/models/DataRetention.d.ts +51 -0
  8. package/dist/esm/models/DataRetention.js +57 -0
  9. package/dist/esm/models/Model.d.ts +6 -0
  10. package/dist/esm/models/Model.js +2 -0
  11. package/dist/esm/models/UpdateDataRetentionRequest.d.ts +32 -0
  12. package/dist/esm/models/UpdateDataRetentionRequest.js +43 -0
  13. package/dist/esm/models/index.d.ts +2 -0
  14. package/dist/esm/models/index.js +2 -0
  15. package/dist/esm/runtime.js +1 -1
  16. package/dist/models/DataRetention.d.ts +51 -0
  17. package/dist/models/DataRetention.js +65 -0
  18. package/dist/models/Model.d.ts +6 -0
  19. package/dist/models/Model.js +2 -0
  20. package/dist/models/UpdateDataRetentionRequest.d.ts +32 -0
  21. package/dist/models/UpdateDataRetentionRequest.js +50 -0
  22. package/dist/models/index.d.ts +2 -0
  23. package/dist/models/index.js +2 -0
  24. package/dist/runtime.js +1 -1
  25. package/docs/DataRetention.md +38 -0
  26. package/docs/FineTuningApi.md +2 -2
  27. package/docs/InferenceApi.md +163 -12
  28. package/docs/Model.md +2 -0
  29. package/docs/UpdateDataRetentionRequest.md +34 -0
  30. package/package.json +1 -1
  31. package/src/apis/InferenceApi.ts +162 -0
  32. package/src/models/DataRetention.ts +95 -0
  33. package/src/models/Model.ts +8 -0
  34. package/src/models/UpdateDataRetentionRequest.ts +66 -0
  35. package/src/models/index.ts +2 -0
  36. package/src/runtime.ts +1 -1
@@ -9,11 +9,13 @@ All URIs are relative to *https://api.gpu.ai/v1*
9
9
  | [**createEmbeddings**](InferenceApi.md#createembeddings) | **POST** /embeddings | Create embeddings (OpenAI-compatible) |
10
10
  | [**createImage**](InferenceApi.md#createimage) | **POST** /images/generations | Create image (OpenAI-compatible) |
11
11
  | [**createVideo**](InferenceApi.md#createvideo) | **POST** /videos | Create a video generation job (async) |
12
+ | [**getDataRetention**](InferenceApi.md#getdataretention) | **GET** /organization/data-retention | Get the org zero data retention setting |
12
13
  | [**getModel**](InferenceApi.md#getmodel) | **GET** /models/{id} | Get a specific model (OpenAI-compatible) |
13
14
  | [**getVideo**](InferenceApi.md#getvideo) | **GET** /videos/{id} | Get a video generation job |
14
15
  | [**getVideoContent**](InferenceApi.md#getvideocontent) | **GET** /videos/{id}/content | Download a completed video artifact |
15
16
  | [**listModels**](InferenceApi.md#listmodels) | **GET** /models | List available models (OpenAI-compatible) |
16
17
  | [**listVideos**](InferenceApi.md#listvideos) | **GET** /videos | List video generation jobs |
18
+ | [**updateDataRetention**](InferenceApi.md#updatedataretentionoperation) | **PUT** /organization/data-retention | Turn zero data retention on or off for the org |
17
19
 
18
20
 
19
21
 
@@ -85,7 +87,7 @@ example().catch(console.error);
85
87
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
86
88
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
87
89
  | **404** | Model or resource not found (OpenAI error envelope). | - |
88
- | **409** | Invalid request (OpenAI error envelope). | - |
90
+ | **409** | Invalid request (OpenAI error envelope). On inference routes this includes `zero_data_retention_unavailable` when zero data retention is required and the model has no zero-data-retention host. | - |
89
91
  | **429** | Rate limit exceeded (OpenAI error envelope). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
90
92
 
91
93
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
@@ -93,7 +95,7 @@ example().catch(console.error);
93
95
 
94
96
  ## createChatCompletion
95
97
 
96
- > ChatCompletionResponse createChatCompletion(chatCompletionRequest)
98
+ > ChatCompletionResponse createChatCompletion(chatCompletionRequest, xGPUAIZeroDataRetention)
97
99
 
98
100
  Create a chat completion (OpenAI-compatible)
99
101
 
@@ -117,6 +119,8 @@ async function example() {
117
119
  const body = {
118
120
  // ChatCompletionRequest
119
121
  chatCompletionRequest: ...,
122
+ // 'true' | 'false' | Send `true` to require zero data retention for this request even when the org setting is off. It can only tighten routing: `false` never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with `true`. A model without a zero-data-retention host returns 400 `zero_data_retention_unavailable`; a value other than true/false returns 400 `invalid_header`. (optional)
123
+ xGPUAIZeroDataRetention: xGPUAIZeroDataRetention_example,
120
124
  } satisfies CreateChatCompletionRequest;
121
125
 
122
126
  try {
@@ -137,6 +141,7 @@ example().catch(console.error);
137
141
  | Name | Type | Description | Notes |
138
142
  |------------- | ------------- | ------------- | -------------|
139
143
  | **chatCompletionRequest** | [ChatCompletionRequest](ChatCompletionRequest.md) | | |
144
+ | **xGPUAIZeroDataRetention** | `true`, `false` | Send &#x60;true&#x60; to require zero data retention for this request even when the org setting is off. It can only tighten routing: &#x60;false&#x60; never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with &#x60;true&#x60;. A model without a zero-data-retention host returns 400 &#x60;zero_data_retention_unavailable&#x60;; a value other than true/false returns 400 &#x60;invalid_header&#x60;. | [Optional] [Defaults to `undefined`] [Enum: true, false] |
140
145
 
141
146
  ### Return type
142
147
 
@@ -156,7 +161,7 @@ example().catch(console.error);
156
161
  | Status code | Description | Response headers |
157
162
  |-------------|-------------|------------------|
158
163
  | **200** | Successful response. application/json for non-streaming requests; text/event-stream for streaming requests (stream&#x3D;true). | - |
159
- | **400** | Invalid request (OpenAI error envelope). | - |
164
+ | **400** | Invalid request (OpenAI error envelope). On inference routes this includes &#x60;zero_data_retention_unavailable&#x60; when zero data retention is required and the model has no zero-data-retention host. | - |
160
165
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
161
166
  | **402** | Insufficient balance (OpenAI error envelope, SERV-06). | - |
162
167
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
@@ -164,14 +169,14 @@ example().catch(console.error);
164
169
  | **422** | Request validation failed (OpenAI error envelope). | - |
165
170
  | **429** | Rate limit exceeded (OpenAI error envelope). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
166
171
  | **500** | Internal server error (OpenAI error envelope). | - |
167
- | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
172
+ | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. Code &#x60;data_retention_check_failed&#x60; means the org zero data retention setting could not be read, so the request was not sent; retry shortly. | * Retry-After - Seconds the client should wait before retrying. <br> |
168
173
 
169
174
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
170
175
 
171
176
 
172
177
  ## createEmbeddings
173
178
 
174
- > EmbeddingsResponse createEmbeddings(embeddingsRequest)
179
+ > EmbeddingsResponse createEmbeddings(embeddingsRequest, xGPUAIZeroDataRetention)
175
180
 
176
181
  Create embeddings (OpenAI-compatible)
177
182
 
@@ -197,6 +202,8 @@ async function example() {
197
202
  const body = {
198
203
  // EmbeddingsRequest
199
204
  embeddingsRequest: ...,
205
+ // 'true' | 'false' | Send `true` to require zero data retention for this request even when the org setting is off. It can only tighten routing: `false` never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with `true`. A model without a zero-data-retention host returns 400 `zero_data_retention_unavailable`; a value other than true/false returns 400 `invalid_header`. (optional)
206
+ xGPUAIZeroDataRetention: xGPUAIZeroDataRetention_example,
200
207
  } satisfies CreateEmbeddingsRequest;
201
208
 
202
209
  try {
@@ -217,6 +224,7 @@ example().catch(console.error);
217
224
  | Name | Type | Description | Notes |
218
225
  |------------- | ------------- | ------------- | -------------|
219
226
  | **embeddingsRequest** | [EmbeddingsRequest](EmbeddingsRequest.md) | | |
227
+ | **xGPUAIZeroDataRetention** | `true`, `false` | Send &#x60;true&#x60; to require zero data retention for this request even when the org setting is off. It can only tighten routing: &#x60;false&#x60; never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with &#x60;true&#x60;. A model without a zero-data-retention host returns 400 &#x60;zero_data_retention_unavailable&#x60;; a value other than true/false returns 400 &#x60;invalid_header&#x60;. | [Optional] [Defaults to `undefined`] [Enum: true, false] |
220
228
 
221
229
  ### Return type
222
230
 
@@ -236,7 +244,7 @@ example().catch(console.error);
236
244
  | Status code | Description | Response headers |
237
245
  |-------------|-------------|------------------|
238
246
  | **200** | One embedding vector per input, in request order. | - |
239
- | **400** | Invalid request (OpenAI error envelope). | - |
247
+ | **400** | Invalid request (OpenAI error envelope). On inference routes this includes &#x60;zero_data_retention_unavailable&#x60; when zero data retention is required and the model has no zero-data-retention host. | - |
240
248
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
241
249
  | **402** | Insufficient balance (OpenAI error envelope, SERV-06). | - |
242
250
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
@@ -245,7 +253,7 @@ example().catch(console.error);
245
253
  | **429** | Rate limit exceeded (OpenAI error envelope). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
246
254
  | **500** | Internal server error (OpenAI error envelope). | - |
247
255
  | **502** | The upstream answered but its response could not be used (OpenAI error envelope): code upstream_response_too_large when the vector payload exceeded the per-request size limit, or upstream_error for a malformed / misaligned response. Not retryable as-is for upstream_response_too_large — send fewer inputs. | - |
248
- | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
256
+ | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. Code &#x60;data_retention_check_failed&#x60; means the org zero data retention setting could not be read, so the request was not sent; retry shortly. | * Retry-After - Seconds the client should wait before retrying. <br> |
249
257
  | **504** | Upstream provider exceeded the synchronous per-attempt deadline (OpenAI error envelope). 504 semantic — distinct from 503 OpenAIUpstreamUnavailable which signals a connection or routing failure rather than a timeout. | - |
250
258
 
251
259
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
@@ -253,7 +261,7 @@ example().catch(console.error);
253
261
 
254
262
  ## createImage
255
263
 
256
- > ImagesResponse createImage(imagesGenerationsRequest, idempotencyKey)
264
+ > ImagesResponse createImage(imagesGenerationsRequest, idempotencyKey, xGPUAIZeroDataRetention)
257
265
 
258
266
  Create image (OpenAI-compatible)
259
267
 
@@ -281,6 +289,8 @@ async function example() {
281
289
  imagesGenerationsRequest: ...,
282
290
  // string (optional)
283
291
  idempotencyKey: idempotencyKey_example,
292
+ // 'true' | 'false' | Send `true` to require zero data retention for this request even when the org setting is off. It can only tighten routing: `false` never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with `true`. A model without a zero-data-retention host returns 400 `zero_data_retention_unavailable`; a value other than true/false returns 400 `invalid_header`. (optional)
293
+ xGPUAIZeroDataRetention: xGPUAIZeroDataRetention_example,
284
294
  } satisfies CreateImageRequest;
285
295
 
286
296
  try {
@@ -302,6 +312,7 @@ example().catch(console.error);
302
312
  |------------- | ------------- | ------------- | -------------|
303
313
  | **imagesGenerationsRequest** | [ImagesGenerationsRequest](ImagesGenerationsRequest.md) | | |
304
314
  | **idempotencyKey** | `string` | | [Optional] [Defaults to `undefined`] |
315
+ | **xGPUAIZeroDataRetention** | `true`, `false` | Send &#x60;true&#x60; to require zero data retention for this request even when the org setting is off. It can only tighten routing: &#x60;false&#x60; never relaxes an org with zero data retention on. Responses routed under zero data retention echo this header with &#x60;true&#x60;. A model without a zero-data-retention host returns 400 &#x60;zero_data_retention_unavailable&#x60;; a value other than true/false returns 400 &#x60;invalid_header&#x60;. | [Optional] [Defaults to `undefined`] [Enum: true, false] |
305
316
 
306
317
  ### Return type
307
318
 
@@ -321,7 +332,7 @@ example().catch(console.error);
321
332
  | Status code | Description | Response headers |
322
333
  |-------------|-------------|------------------|
323
334
  | **200** | OpenAI-shaped image generation response (b64_json only). | - |
324
- | **400** | Invalid request (OpenAI error envelope). | - |
335
+ | **400** | Invalid request (OpenAI error envelope). On inference routes this includes &#x60;zero_data_retention_unavailable&#x60; when zero data retention is required and the model has no zero-data-retention host. | - |
325
336
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
326
337
  | **402** | Insufficient balance (OpenAI error envelope, SERV-06). | - |
327
338
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
@@ -329,7 +340,7 @@ example().catch(console.error);
329
340
  | **422** | Request validation failed (OpenAI error envelope). | - |
330
341
  | **429** | Rate limit exceeded (OpenAI error envelope). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
331
342
  | **500** | Internal server error (OpenAI error envelope). | - |
332
- | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
343
+ | **503** | Upstream provider unavailable (OpenAI error envelope). Retry-After header may indicate seconds to wait. Code &#x60;data_retention_check_failed&#x60; means the org zero data retention setting could not be read, so the request was not sent; retry shortly. | * Retry-After - Seconds the client should wait before retrying. <br> |
333
344
  | **504** | Upstream provider exceeded the synchronous per-attempt deadline (OpenAI error envelope). 504 semantic — distinct from 503 OpenAIUpstreamUnavailable which signals a connection or routing failure rather than a timeout. | - |
334
345
 
335
346
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
@@ -405,7 +416,7 @@ example().catch(console.error);
405
416
  | Status code | Description | Response headers |
406
417
  |-------------|-------------|------------------|
407
418
  | **200** | A video generation job object. | - |
408
- | **400** | Invalid request (OpenAI error envelope). | - |
419
+ | **400** | Invalid request (OpenAI error envelope). On inference routes this includes &#x60;zero_data_retention_unavailable&#x60; when zero data retention is required and the model has no zero-data-retention host. | - |
409
420
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
410
421
  | **402** | Insufficient balance (OpenAI error envelope, SERV-06). | - |
411
422
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
@@ -416,6 +427,71 @@ example().catch(console.error);
416
427
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
417
428
 
418
429
 
430
+ ## getDataRetention
431
+
432
+ > DataRetention getDataRetention()
433
+
434
+ Get the org zero data retention setting
435
+
436
+ Returns whether zero data retention (ZDR) is on for the organization. When it is on, chat, embeddings and image requests route only to model hosts that keep no prompt or output content, a model without one returns 400 &#x60;zero_data_retention_unavailable&#x60;, and video generation is refused. Requires the &#x60;serverless:read&#x60; scope. The organization is taken from the authenticated API key.
437
+
438
+ ### Example
439
+
440
+ ```ts
441
+ import {
442
+ Configuration,
443
+ InferenceApi,
444
+ } from '@gpuai/sdk';
445
+ import type { GetDataRetentionRequest } from '@gpuai/sdk';
446
+
447
+ async function example() {
448
+ console.log("🚀 Testing @gpuai/sdk SDK...");
449
+ const config = new Configuration({
450
+ // Configure HTTP bearer authorization: bearerAuth
451
+ accessToken: "YOUR BEARER TOKEN",
452
+ });
453
+ const api = new InferenceApi(config);
454
+
455
+ try {
456
+ const data = await api.getDataRetention();
457
+ console.log(data);
458
+ } catch (error) {
459
+ console.error(error);
460
+ }
461
+ }
462
+
463
+ // Run the test
464
+ example().catch(console.error);
465
+ ```
466
+
467
+ ### Parameters
468
+
469
+ This endpoint does not need any parameter.
470
+
471
+ ### Return type
472
+
473
+ [**DataRetention**](DataRetention.md)
474
+
475
+ ### Authorization
476
+
477
+ [bearerAuth](../README.md#bearerAuth)
478
+
479
+ ### HTTP request headers
480
+
481
+ - **Content-Type**: Not defined
482
+ - **Accept**: `application/json`, `application/problem+json`
483
+
484
+
485
+ ### HTTP response details
486
+ | Status code | Description | Response headers |
487
+ |-------------|-------------|------------------|
488
+ | **200** | OK | - |
489
+ | **429** | Rate limit exceeded (RFC 7807). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
490
+ | **0** | Error response (RFC 7807) | - |
491
+
492
+ [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
493
+
494
+
419
495
  ## getModel
420
496
 
421
497
  > Model getModel(id)
@@ -703,7 +779,7 @@ example().catch(console.error);
703
779
  | Status code | Description | Response headers |
704
780
  |-------------|-------------|------------------|
705
781
  | **200** | OpenAI-shaped model list. | - |
706
- | **400** | Invalid request (OpenAI error envelope). | - |
782
+ | **400** | Invalid request (OpenAI error envelope). On inference routes this includes &#x60;zero_data_retention_unavailable&#x60; when zero data retention is required and the model has no zero-data-retention host. | - |
707
783
  | **401** | Missing or invalid API key (OpenAI error envelope). | - |
708
784
  | **403** | API key lacks the required scope (OpenAI error envelope). | - |
709
785
  | **429** | Rate limit exceeded (OpenAI error envelope). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
@@ -785,3 +861,78 @@ example().catch(console.error);
785
861
 
786
862
  [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
787
863
 
864
+
865
+ ## updateDataRetention
866
+
867
+ > DataRetention updateDataRetention(updateDataRetentionRequest)
868
+
869
+ Turn zero data retention on or off for the org
870
+
871
+ Sets the organization\&#39;s zero data retention (ZDR) setting. Takes effect from the next inference request. Requires the &#x60;serverless:write&#x60; scope AND org-admin privileges (a non-admin member gets 403). &#x60;zero_data_retention&#x60; is required (422 when missing).
872
+
873
+ ### Example
874
+
875
+ ```ts
876
+ import {
877
+ Configuration,
878
+ InferenceApi,
879
+ } from '@gpuai/sdk';
880
+ import type { UpdateDataRetentionOperationRequest } from '@gpuai/sdk';
881
+
882
+ async function example() {
883
+ console.log("🚀 Testing @gpuai/sdk SDK...");
884
+ const config = new Configuration({
885
+ // Configure HTTP bearer authorization: bearerAuth
886
+ accessToken: "YOUR BEARER TOKEN",
887
+ });
888
+ const api = new InferenceApi(config);
889
+
890
+ const body = {
891
+ // UpdateDataRetentionRequest
892
+ updateDataRetentionRequest: ...,
893
+ } satisfies UpdateDataRetentionOperationRequest;
894
+
895
+ try {
896
+ const data = await api.updateDataRetention(body);
897
+ console.log(data);
898
+ } catch (error) {
899
+ console.error(error);
900
+ }
901
+ }
902
+
903
+ // Run the test
904
+ example().catch(console.error);
905
+ ```
906
+
907
+ ### Parameters
908
+
909
+
910
+ | Name | Type | Description | Notes |
911
+ |------------- | ------------- | ------------- | -------------|
912
+ | **updateDataRetentionRequest** | [UpdateDataRetentionRequest](UpdateDataRetentionRequest.md) | | |
913
+
914
+ ### Return type
915
+
916
+ [**DataRetention**](DataRetention.md)
917
+
918
+ ### Authorization
919
+
920
+ [bearerAuth](../README.md#bearerAuth)
921
+
922
+ ### HTTP request headers
923
+
924
+ - **Content-Type**: `application/json`
925
+ - **Accept**: `application/json`, `application/problem+json`
926
+
927
+
928
+ ### HTTP response details
929
+ | Status code | Description | Response headers |
930
+ |-------------|-------------|------------------|
931
+ | **200** | OK | - |
932
+ | **403** | Error response (RFC 7807) | - |
933
+ | **422** | Error response (RFC 7807) | - |
934
+ | **429** | Rate limit exceeded (RFC 7807). Retry-After header indicates seconds to wait. | * Retry-After - Seconds the client should wait before retrying. <br> |
935
+ | **0** | Error response (RFC 7807) | - |
936
+
937
+ [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
938
+
package/docs/Model.md CHANGED
@@ -17,6 +17,7 @@ Name | Type
17
17
  `supportedParameters` | Array&lt;string&gt;
18
18
  `supportsTools` | boolean
19
19
  `supportsStructuredOutputs` | boolean
20
+ `zeroDataRetention` | boolean
20
21
  `parameters` | [Array&lt;ModelParametersInner&gt;](ModelParametersInner.md)
21
22
  `pricing` | [ModelPricing](ModelPricing.md)
22
23
  `aliases` | Array&lt;string&gt;
@@ -42,6 +43,7 @@ const example = {
42
43
  "supportedParameters": null,
43
44
  "supportsTools": null,
44
45
  "supportsStructuredOutputs": null,
46
+ "zeroDataRetention": null,
45
47
  "parameters": null,
46
48
  "pricing": null,
47
49
  "aliases": null,
@@ -0,0 +1,34 @@
1
+
2
+ # UpdateDataRetentionRequest
3
+
4
+
5
+ ## Properties
6
+
7
+ Name | Type
8
+ ------------ | -------------
9
+ `zeroDataRetention` | boolean
10
+
11
+ ## Example
12
+
13
+ ```typescript
14
+ import type { UpdateDataRetentionRequest } from '@gpuai/sdk'
15
+
16
+ // TODO: Update the object below with actual values
17
+ const example = {
18
+ "zeroDataRetention": null,
19
+ } satisfies UpdateDataRetentionRequest
20
+
21
+ console.log(example)
22
+
23
+ // Convert the instance to a JSON string
24
+ const exampleJSON: string = JSON.stringify(example)
25
+ console.log(exampleJSON)
26
+
27
+ // Parse the JSON string back to an object
28
+ const exampleParsed = JSON.parse(exampleJSON) as UpdateDataRetentionRequest
29
+ console.log(exampleParsed)
30
+ ```
31
+
32
+ [[Back to top]](#) [[Back to API list]](../README.md#api-endpoints) [[Back to Model list]](../README.md#models) [[Back to README]](../README.md)
33
+
34
+
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@gpuai/sdk",
3
- "version": "0.5.1",
3
+ "version": "0.5.3",
4
4
  "description": "TypeScript SDK for the GPU.ai Public Developer API",
5
5
  "author": "GPU.ai Support <support@gpu.ai>",
6
6
  "repository": {
@@ -23,6 +23,11 @@ import {
23
23
  ChatCompletionResponseFromJSON,
24
24
  ChatCompletionResponseToJSON,
25
25
  } from '../models/ChatCompletionResponse';
26
+ import {
27
+ type DataRetention,
28
+ DataRetentionFromJSON,
29
+ DataRetentionToJSON,
30
+ } from '../models/DataRetention';
26
31
  import {
27
32
  type EmbeddingsRequest,
28
33
  EmbeddingsRequestFromJSON,
@@ -58,6 +63,16 @@ import {
58
63
  OpenAIErrorEnvelopeFromJSON,
59
64
  OpenAIErrorEnvelopeToJSON,
60
65
  } from '../models/OpenAIErrorEnvelope';
66
+ import {
67
+ type Problem,
68
+ ProblemFromJSON,
69
+ ProblemToJSON,
70
+ } from '../models/Problem';
71
+ import {
72
+ type UpdateDataRetentionRequest,
73
+ UpdateDataRetentionRequestFromJSON,
74
+ UpdateDataRetentionRequestToJSON,
75
+ } from '../models/UpdateDataRetentionRequest';
61
76
  import {
62
77
  type VideoCreateRequest,
63
78
  VideoCreateRequestFromJSON,
@@ -80,15 +95,18 @@ export interface CancelVideoRequest {
80
95
 
81
96
  export interface CreateChatCompletionRequest {
82
97
  chatCompletionRequest: ChatCompletionRequest;
98
+ xGPUAIZeroDataRetention?: CreateChatCompletionXGPUAIZeroDataRetentionEnum;
83
99
  }
84
100
 
85
101
  export interface CreateEmbeddingsRequest {
86
102
  embeddingsRequest: EmbeddingsRequest;
103
+ xGPUAIZeroDataRetention?: CreateEmbeddingsXGPUAIZeroDataRetentionEnum;
87
104
  }
88
105
 
89
106
  export interface CreateImageRequest {
90
107
  imagesGenerationsRequest: ImagesGenerationsRequest;
91
108
  idempotencyKey?: string;
109
+ xGPUAIZeroDataRetention?: CreateImageXGPUAIZeroDataRetentionEnum;
92
110
  }
93
111
 
94
112
  export interface CreateVideoRequest {
@@ -117,6 +135,10 @@ export interface ListVideosRequest {
117
135
  limit?: number;
118
136
  }
119
137
 
138
+ export interface UpdateDataRetentionOperationRequest {
139
+ updateDataRetentionRequest: UpdateDataRetentionRequest;
140
+ }
141
+
120
142
  /**
121
143
  *
122
144
  */
@@ -192,6 +214,10 @@ export class InferenceApi extends runtime.BaseAPI {
192
214
 
193
215
  headerParameters['Content-Type'] = 'application/json';
194
216
 
217
+ if (requestParameters['xGPUAIZeroDataRetention'] != null) {
218
+ headerParameters['X-GPUAI-Zero-Data-Retention'] = String(requestParameters['xGPUAIZeroDataRetention']);
219
+ }
220
+
195
221
  if (this.configuration && this.configuration.accessToken) {
196
222
  const token = this.configuration.accessToken;
197
223
  const tokenString = await token("bearerAuth", []);
@@ -247,6 +273,10 @@ export class InferenceApi extends runtime.BaseAPI {
247
273
 
248
274
  headerParameters['Content-Type'] = 'application/json';
249
275
 
276
+ if (requestParameters['xGPUAIZeroDataRetention'] != null) {
277
+ headerParameters['X-GPUAI-Zero-Data-Retention'] = String(requestParameters['xGPUAIZeroDataRetention']);
278
+ }
279
+
250
280
  if (this.configuration && this.configuration.accessToken) {
251
281
  const token = this.configuration.accessToken;
252
282
  const tokenString = await token("bearerAuth", []);
@@ -308,6 +338,10 @@ export class InferenceApi extends runtime.BaseAPI {
308
338
  headerParameters['Idempotency-Key'] = String(requestParameters['idempotencyKey']);
309
339
  }
310
340
 
341
+ if (requestParameters['xGPUAIZeroDataRetention'] != null) {
342
+ headerParameters['X-GPUAI-Zero-Data-Retention'] = String(requestParameters['xGPUAIZeroDataRetention']);
343
+ }
344
+
311
345
  if (this.configuration && this.configuration.accessToken) {
312
346
  const token = this.configuration.accessToken;
313
347
  const tokenString = await token("bearerAuth", []);
@@ -409,6 +443,53 @@ export class InferenceApi extends runtime.BaseAPI {
409
443
  return await response.value();
410
444
  }
411
445
 
446
+ /**
447
+ * Creates request options for getDataRetention without sending the request
448
+ */
449
+ async getDataRetentionRequestOpts(): Promise<runtime.RequestOpts> {
450
+ const queryParameters: any = {};
451
+
452
+ const headerParameters: runtime.HTTPHeaders = {};
453
+
454
+ if (this.configuration && this.configuration.accessToken) {
455
+ const token = this.configuration.accessToken;
456
+ const tokenString = await token("bearerAuth", []);
457
+
458
+ if (tokenString) {
459
+ headerParameters["Authorization"] = `Bearer ${tokenString}`;
460
+ }
461
+ }
462
+
463
+ let urlPath = `/organization/data-retention`;
464
+
465
+ return {
466
+ path: urlPath,
467
+ method: 'GET',
468
+ headers: headerParameters,
469
+ query: queryParameters,
470
+ };
471
+ }
472
+
473
+ /**
474
+ * Returns whether zero data retention (ZDR) is on for the organization. When it is on, chat, embeddings and image requests route only to model hosts that keep no prompt or output content, a model without one returns 400 `zero_data_retention_unavailable`, and video generation is refused. Requires the `serverless:read` scope. The organization is taken from the authenticated API key.
475
+ * Get the org zero data retention setting
476
+ */
477
+ async getDataRetentionRaw(initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<DataRetention>> {
478
+ const requestOptions = await this.getDataRetentionRequestOpts();
479
+ const response = await this.request(requestOptions, initOverrides);
480
+
481
+ return new runtime.JSONApiResponse(response, (jsonValue) => DataRetentionFromJSON(jsonValue));
482
+ }
483
+
484
+ /**
485
+ * Returns whether zero data retention (ZDR) is on for the organization. When it is on, chat, embeddings and image requests route only to model hosts that keep no prompt or output content, a model without one returns 400 `zero_data_retention_unavailable`, and video generation is refused. Requires the `serverless:read` scope. The organization is taken from the authenticated API key.
486
+ * Get the org zero data retention setting
487
+ */
488
+ async getDataRetention(initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<DataRetention> {
489
+ const response = await this.getDataRetentionRaw(initOverrides);
490
+ return await response.value();
491
+ }
492
+
412
493
  /**
413
494
  * Creates request options for getModel without sending the request
414
495
  */
@@ -672,8 +753,89 @@ export class InferenceApi extends runtime.BaseAPI {
672
753
  return await response.value();
673
754
  }
674
755
 
756
+ /**
757
+ * Creates request options for updateDataRetention without sending the request
758
+ */
759
+ async updateDataRetentionRequestOpts(requestParameters: UpdateDataRetentionOperationRequest): Promise<runtime.RequestOpts> {
760
+ if (requestParameters['updateDataRetentionRequest'] == null) {
761
+ throw new runtime.RequiredError(
762
+ 'updateDataRetentionRequest',
763
+ 'Required parameter "updateDataRetentionRequest" was null or undefined when calling updateDataRetention().'
764
+ );
765
+ }
766
+
767
+ const queryParameters: any = {};
768
+
769
+ const headerParameters: runtime.HTTPHeaders = {};
770
+
771
+ headerParameters['Content-Type'] = 'application/json';
772
+
773
+ if (this.configuration && this.configuration.accessToken) {
774
+ const token = this.configuration.accessToken;
775
+ const tokenString = await token("bearerAuth", []);
776
+
777
+ if (tokenString) {
778
+ headerParameters["Authorization"] = `Bearer ${tokenString}`;
779
+ }
780
+ }
781
+
782
+ let urlPath = `/organization/data-retention`;
783
+
784
+ return {
785
+ path: urlPath,
786
+ method: 'PUT',
787
+ headers: headerParameters,
788
+ query: queryParameters,
789
+ body: UpdateDataRetentionRequestToJSON(requestParameters['updateDataRetentionRequest']),
790
+ };
791
+ }
792
+
793
+ /**
794
+ * Sets the organization\'s zero data retention (ZDR) setting. Takes effect from the next inference request. Requires the `serverless:write` scope AND org-admin privileges (a non-admin member gets 403). `zero_data_retention` is required (422 when missing).
795
+ * Turn zero data retention on or off for the org
796
+ */
797
+ async updateDataRetentionRaw(requestParameters: UpdateDataRetentionOperationRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<runtime.ApiResponse<DataRetention>> {
798
+ const requestOptions = await this.updateDataRetentionRequestOpts(requestParameters);
799
+ const response = await this.request(requestOptions, initOverrides);
800
+
801
+ return new runtime.JSONApiResponse(response, (jsonValue) => DataRetentionFromJSON(jsonValue));
802
+ }
803
+
804
+ /**
805
+ * Sets the organization\'s zero data retention (ZDR) setting. Takes effect from the next inference request. Requires the `serverless:write` scope AND org-admin privileges (a non-admin member gets 403). `zero_data_retention` is required (422 when missing).
806
+ * Turn zero data retention on or off for the org
807
+ */
808
+ async updateDataRetention(requestParameters: UpdateDataRetentionOperationRequest, initOverrides?: RequestInit | runtime.InitOverrideFunction): Promise<DataRetention> {
809
+ const response = await this.updateDataRetentionRaw(requestParameters, initOverrides);
810
+ return await response.value();
811
+ }
812
+
675
813
  }
676
814
 
815
+ /**
816
+ * @export
817
+ */
818
+ export const CreateChatCompletionXGPUAIZeroDataRetentionEnum = {
819
+ True: 'true',
820
+ False: 'false'
821
+ } as const;
822
+ export type CreateChatCompletionXGPUAIZeroDataRetentionEnum = typeof CreateChatCompletionXGPUAIZeroDataRetentionEnum[keyof typeof CreateChatCompletionXGPUAIZeroDataRetentionEnum];
823
+ /**
824
+ * @export
825
+ */
826
+ export const CreateEmbeddingsXGPUAIZeroDataRetentionEnum = {
827
+ True: 'true',
828
+ False: 'false'
829
+ } as const;
830
+ export type CreateEmbeddingsXGPUAIZeroDataRetentionEnum = typeof CreateEmbeddingsXGPUAIZeroDataRetentionEnum[keyof typeof CreateEmbeddingsXGPUAIZeroDataRetentionEnum];
831
+ /**
832
+ * @export
833
+ */
834
+ export const CreateImageXGPUAIZeroDataRetentionEnum = {
835
+ True: 'true',
836
+ False: 'false'
837
+ } as const;
838
+ export type CreateImageXGPUAIZeroDataRetentionEnum = typeof CreateImageXGPUAIZeroDataRetentionEnum[keyof typeof CreateImageXGPUAIZeroDataRetentionEnum];
677
839
  /**
678
840
  * @export
679
841
  */