java/resources/conversations/methods/create/index.md 2026-09-16 22:58 UTC to 2026-09-18 22:59 UTC

1## Create a conversation

2 

3`Conversation conversations().create(ConversationCreateParamsparams = ConversationCreateParams.none(), RequestOptionsrequestOptions = RequestOptions.none())`

4 

5**post** `/conversations`

6 

7Create a conversation.

8 

9### Parameters

10 

11- `ConversationCreateParams params`

12 

13 - `Optional<List<ResponseInputItem>> items`

14 

15 Initial items to include in the conversation context. You may add up to 20 items at a time.

16 

17 - `class EasyInputMessage:`

18 

19 A message input to the model with a role indicating instruction following

20 hierarchy. Instructions given with the `developer` or `system` role take

21 precedence over instructions given with the `user` role. Messages with the

22 `assistant` role are presumed to have been generated by the model in previous

23 interactions.

24 

25 - `Content content`

26 

27 Text, image, or audio input to the model, used to generate a response.

28 Can also contain previous assistant responses.

29 

30 - `String`

31 

32 - `List<ResponseInputContent>`

33 

34 - `class ResponseInputText:`

35 

36 A text input to the model.

37 

38 - `String text`

39 

40 The text input to the model.

41 

42 - `JsonValue; type "input_text"constant`

43 

44 The type of the input item. Always `input_text`.

45 

46 - `INPUT_TEXT("input_text")`

47 

48 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

49 

50 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

51 

52 - `JsonValue; mode "explicit"constant`

53 

54 The breakpoint mode. Always `explicit`.

55 

56 - `EXPLICIT("explicit")`

57 

58 - `class ResponseInputImage:`

59 

60 An image input to the model. Learn about [image inputs](https://platform.openai.com/docs/guides/vision).

61 

62 - `Detail detail`

63 

64 The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`.

65 

66 - `LOW("low")`

67 

68 - `HIGH("high")`

69 

70 - `AUTO("auto")`

71 

72 - `ORIGINAL("original")`

73 

74 - `JsonValue; type "input_image"constant`

75 

76 The type of the input item. Always `input_image`.

77 

78 - `INPUT_IMAGE("input_image")`

79 

80 - `Optional<String> fileId`

81 

82 The ID of the file to be sent to the model.

83 

84 - `Optional<String> imageUrl`

85 

86 The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL.

87 

88 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

89 

90 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

91 

92 - `JsonValue; mode "explicit"constant`

93 

94 The breakpoint mode. Always `explicit`.

95 

96 - `EXPLICIT("explicit")`

97 

98 - `class ResponseInputFile:`

99 

100 A file input to the model.

101 

102 - `JsonValue; type "input_file"constant`

103 

104 The type of the input item. Always `input_file`.

105 

106 - `INPUT_FILE("input_file")`

107 

108 - `Optional<Detail> detail`

109 

110 The detail level of the file to be sent to the model. Use `auto` to let the system select the detail level; for GPT-5.6 and later models, `auto` uses high-quality rendering, which may increase input token usage. Use `low` for lower-cost rendering, or `high` to render the file at higher quality. Defaults to `auto`.

111 

112 - `AUTO("auto")`

113 

114 - `LOW("low")`

115 

116 - `HIGH("high")`

117 

118 - `Optional<String> fileData`

119 

120 The content of the file to be sent to the model.

121 

122 - `Optional<String> fileId`

123 

124 The ID of the file to be sent to the model.

125 

126 - `Optional<String> fileUrl`

127 

128 The URL of the file to be sent to the model.

129 

130 - `Optional<String> filename`

131 

132 The name of the file to be sent to the model.

133 

134 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

135 

136 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

137 

138 - `JsonValue; mode "explicit"constant`

139 

140 The breakpoint mode. Always `explicit`.

141 

142 - `EXPLICIT("explicit")`

143 

144 - `Role role`

145 

146 The role of the message input. One of `user`, `assistant`, `system`, or

147 `developer`.

148 

149 - `USER("user")`

150 

151 - `ASSISTANT("assistant")`

152 

153 - `SYSTEM("system")`

154 

155 - `DEVELOPER("developer")`

156 

157 - `Optional<Phase> phase`

158 

159 Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`).

160 For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend

161 phase on all assistant messages — dropping it can degrade performance. Not used for user messages.

162 

163 - `COMMENTARY("commentary")`

164 

165 - `FINAL_ANSWER("final_answer")`

166 

167 - `Optional<Type> type`

168 

169 The type of the message input. Always `message`.

170 

171 - `MESSAGE("message")`

172 

173 - `Message`

174 

175 - `List<ResponseInputContent> content`

176 

177 A list of one or many input items to the model, containing different content

178 types.

179 

180 - `class ResponseInputText:`

181 

182 A text input to the model.

183 

184 - `class ResponseInputImage:`

185 

186 An image input to the model. Learn about [image inputs](https://platform.openai.com/docs/guides/vision).

187 

188 - `class ResponseInputFile:`

189 

190 A file input to the model.

191 

192 - `Role role`

193 

194 The role of the message input. One of `user`, `system`, or `developer`.

195 

196 - `USER("user")`

197 

198 - `SYSTEM("system")`

199 

200 - `DEVELOPER("developer")`

201 

202 - `Optional<Status> status`

203 

204 The status of item. One of `in_progress`, `completed`, or

205 `incomplete`. Populated when items are returned via API.

206 

207 - `IN_PROGRESS("in_progress")`

208 

209 - `COMPLETED("completed")`

210 

211 - `INCOMPLETE("incomplete")`

212 

213 - `Optional<Type> type`

214 

215 The type of the message input. Always set to `message`.

216 

217 - `MESSAGE("message")`

218 

219 - `class ResponseOutputMessage:`

220 

221 An output message from the model.

222 

223 - `String id`

224 

225 The unique ID of the output message.

226 

227 - `List<Content> content`

228 

229 The content of the output message.

230 

231 - `class ResponseOutputText:`

232 

233 A text output from the model.

234 

235 - `List<Annotation> annotations`

236 

237 The annotations of the text output.

238 

239 - `class FileCitation:`

240 

241 A citation to a file.

242 

243 - `String fileId`

244 

245 The ID of the file.

246 

247 - `String filename`

248 

249 The filename of the file cited.

250 

251 - `long index`

252 

253 The index of the file in the list of files.

254 

255 - `JsonValue; type "file_citation"constant`

256 

257 The type of the file citation. Always `file_citation`.

258 

259 - `FILE_CITATION("file_citation")`

260 

261 - `class UrlCitation:`

262 

263 A citation for a web resource used to generate a model response.

264 

265 - `long endIndex`

266 

267 The index of the last character of the URL citation in the message.

268 

269 - `long startIndex`

270 

271 The index of the first character of the URL citation in the message.

272 

273 - `String title`

274 

275 The title of the web resource.

276 

277 - `JsonValue; type "url_citation"constant`

278 

279 The type of the URL citation. Always `url_citation`.

280 

281 - `URL_CITATION("url_citation")`

282 

283 - `String url`

284 

285 The URL of the web resource.

286 

287 - `class ContainerFileCitation:`

288 

289 A citation for a container file used to generate a model response.

290 

291 - `String containerId`

292 

293 The ID of the container file.

294 

295 - `long endIndex`

296 

297 The index of the last character of the container file citation in the message.

298 

299 - `String fileId`

300 

301 The ID of the file.

302 

303 - `String filename`

304 

305 The filename of the container file cited.

306 

307 - `long startIndex`

308 

309 The index of the first character of the container file citation in the message.

310 

311 - `JsonValue; type "container_file_citation"constant`

312 

313 The type of the container file citation. Always `container_file_citation`.

314 

315 - `CONTAINER_FILE_CITATION("container_file_citation")`

316 

317 - `class FilePath:`

318 

319 A path to a file.

320 

321 - `String fileId`

322 

323 The ID of the file.

324 

325 - `long index`

326 

327 The index of the file in the list of files.

328 

329 - `JsonValue; type "file_path"constant`

330 

331 The type of the file path. Always `file_path`.

332 

333 - `FILE_PATH("file_path")`

334 

335 - `String text`

336 

337 The text output from the model.

338 

339 - `JsonValue; type "output_text"constant`

340 

341 The type of the output text. Always `output_text`.

342 

343 - `OUTPUT_TEXT("output_text")`

344 

345 - `Optional<List<Logprob>> logprobs`

346 

347 - `String token`

348 

349 - `List<long> bytes`

350 

351 - `double logprob`

352 

353 - `List<TopLogprob> topLogprobs`

354 

355 - `String token`

356 

357 - `List<long> bytes`

358 

359 - `double logprob`

360 

361 - `class ResponseOutputRefusal:`

362 

363 A refusal from the model.

364 

365 - `String refusal`

366 

367 The refusal explanation from the model.

368 

369 - `JsonValue; type "refusal"constant`

370 

371 The type of the refusal. Always `refusal`.

372 

373 - `REFUSAL("refusal")`

374 

375 - `JsonValue; role "assistant"constant`

376 

377 The role of the output message. Always `assistant`.

378 

379 - `ASSISTANT("assistant")`

380 

381 - `Status status`

382 

383 The status of the message input. One of `in_progress`, `completed`, or

384 `incomplete`. Populated when input items are returned via API.

385 

386 - `IN_PROGRESS("in_progress")`

387 

388 - `COMPLETED("completed")`

389 

390 - `INCOMPLETE("incomplete")`

391 

392 - `JsonValue; type "message"constant`

393 

394 The type of the output message. Always `message`.

395 

396 - `MESSAGE("message")`

397 

398 - `Optional<Phase> phase`

399 

400 Labels an `assistant` message as intermediate commentary (`commentary`) or the final answer (`final_answer`).

401 For models like `gpt-5.3-codex` and beyond, when sending follow-up requests, preserve and resend

402 phase on all assistant messages — dropping it can degrade performance. Not used for user messages.

403 

404 - `COMMENTARY("commentary")`

405 

406 - `FINAL_ANSWER("final_answer")`

407 

408 - `class ResponseFileSearchToolCall:`

409 

410 The results of a file search tool call. See the

411 [file search guide](https://platform.openai.com/docs/guides/tools-file-search) for more information.

412 

413 - `String id`

414 

415 The unique ID of the file search tool call.

416 

417 - `List<String> queries`

418 

419 The queries used to search for files.

420 

421 - `Status status`

422 

423 The status of the file search tool call. One of `in_progress`,

424 `searching`, `incomplete` or `failed`,

425 

426 - `IN_PROGRESS("in_progress")`

427 

428 - `SEARCHING("searching")`

429 

430 - `COMPLETED("completed")`

431 

432 - `INCOMPLETE("incomplete")`

433 

434 - `FAILED("failed")`

435 

436 - `JsonValue; type "file_search_call"constant`

437 

438 The type of the file search tool call. Always `file_search_call`.

439 

440 - `FILE_SEARCH_CALL("file_search_call")`

441 

442 - `Optional<List<Result>> results`

443 

444 The results of the file search tool call.

445 

446 - `Optional<Attributes> attributes`

447 

448 Set of 16 key-value pairs that can be attached to an object. This can be

449 useful for storing additional information about the object in a structured

450 format, and querying for objects via API or the dashboard. Keys are strings

451 with a maximum length of 64 characters. Values are strings with a maximum

452 length of 512 characters, booleans, or numbers.

453 

454 - `String`

455 

456 - `double`

457 

458 - `boolean`

459 

460 - `Optional<String> fileId`

461 

462 The unique ID of the file.

463 

464 - `Optional<String> filename`

465 

466 The name of the file.

467 

468 - `Optional<Double> score`

469 

470 The relevance score of the file - a value between 0 and 1.

471 

472 - `Optional<String> text`

473 

474 The text that was retrieved from the file.

475 

476 - `class ResponseComputerToolCall:`

477 

478 A tool call to a computer use tool. See the

479 [computer use guide](https://platform.openai.com/docs/guides/tools-computer-use) for more information.

480 

481 - `String id`

482 

483 The unique ID of the computer call.

484 

485 - `String callId`

486 

487 An identifier used when responding to the tool call with output.

488 

489 - `List<PendingSafetyCheck> pendingSafetyChecks`

490 

491 The pending safety checks for the computer call.

492 

493 - `String id`

494 

495 The ID of the pending safety check.

496 

497 - `Optional<String> code`

498 

499 The type of the pending safety check.

500 

501 - `Optional<String> message`

502 

503 Details about the pending safety check.

504 

505 - `Status status`

506 

507 The status of the item. One of `in_progress`, `completed`, or

508 `incomplete`. Populated when items are returned via API.

509 

510 - `IN_PROGRESS("in_progress")`

511 

512 - `COMPLETED("completed")`

513 

514 - `INCOMPLETE("incomplete")`

515 

516 - `Type type`

517 

518 The type of the computer call. Always `computer_call`.

519 

520 - `COMPUTER_CALL("computer_call")`

521 

522 - `Optional<Action> action`

523 

524 A click action.

525 

526 - `class Click:`

527 

528 A click action.

529 

530 - `Button button`

531 

532 Indicates which mouse button was pressed during the click. One of `left`, `right`, `wheel`, `back`, or `forward`.

533 

534 - `LEFT("left")`

535 

536 - `RIGHT("right")`

537 

538 - `WHEEL("wheel")`

539 

540 - `BACK("back")`

541 

542 - `FORWARD("forward")`

543 

544 - `JsonValue; type "click"constant`

545 

546 Specifies the event type. For a click action, this property is always `click`.

547 

548 - `CLICK("click")`

549 

550 - `long x`

551 

552 The x-coordinate where the click occurred.

553 

554 - `long y`

555 

556 The y-coordinate where the click occurred.

557 

558 - `Optional<List<String>> keys`

559 

560 The keys being held while clicking.

561 

562 - `class DoubleClick:`

563 

564 A double click action.

565 

566 - `Optional<List<String>> keys`

567 

568 The keys being held while double-clicking.

569 

570 - `JsonValue; type "double_click"constant`

571 

572 Specifies the event type. For a double click action, this property is always set to `double_click`.

573 

574 - `DOUBLE_CLICK("double_click")`

575 

576 - `long x`

577 

578 The x-coordinate where the double click occurred.

579 

580 - `long y`

581 

582 The y-coordinate where the double click occurred.

583 

584 - `class Drag:`

585 

586 A drag action.

587 

588 - `List<Path> path`

589 

590 An array of coordinates representing the path of the drag action. Coordinates will appear as an array of objects, eg

591 

592 ```

593 [

594 { x: 100, y: 200 },

595 { x: 200, y: 300 }

596 ]

597 ```

598 

599 - `long x`

600 

601 The x-coordinate.

602 

603 - `long y`

604 

605 The y-coordinate.

606 

607 - `JsonValue; type "drag"constant`

608 

609 Specifies the event type. For a drag action, this property is always set to `drag`.

610 

611 - `DRAG("drag")`

612 

613 - `Optional<List<String>> keys`

614 

615 The keys being held while dragging the mouse.

616 

617 - `class Keypress:`

618 

619 A collection of keypresses the model would like to perform.

620 

621 - `List<String> keys`

622 

623 The combination of keys the model is requesting to be pressed. This is an array of strings, each representing a key.

624 

625 - `JsonValue; type "keypress"constant`

626 

627 Specifies the event type. For a keypress action, this property is always set to `keypress`.

628 

629 - `KEYPRESS("keypress")`

630 

631 - `class Move:`

632 

633 A mouse move action.

634 

635 - `JsonValue; type "move"constant`

636 

637 Specifies the event type. For a move action, this property is always set to `move`.

638 

639 - `MOVE("move")`

640 

641 - `long x`

642 

643 The x-coordinate to move to.

644 

645 - `long y`

646 

647 The y-coordinate to move to.

648 

649 - `Optional<List<String>> keys`

650 

651 The keys being held while moving the mouse.

652 

653 - `JsonValue;`

654 

655 - `JsonValue; type "screenshot"constant`

656 

657 Specifies the event type. For a screenshot action, this property is always set to `screenshot`.

658 

659 - `SCREENSHOT("screenshot")`

660 

661 - `class Scroll:`

662 

663 A scroll action.

664 

665 - `long scrollX`

666 

667 The horizontal scroll distance.

668 

669 - `long scrollY`

670 

671 The vertical scroll distance.

672 

673 - `JsonValue; type "scroll"constant`

674 

675 Specifies the event type. For a scroll action, this property is always set to `scroll`.

676 

677 - `SCROLL("scroll")`

678 

679 - `long x`

680 

681 The x-coordinate where the scroll occurred.

682 

683 - `long y`

684 

685 The y-coordinate where the scroll occurred.

686 

687 - `Optional<List<String>> keys`

688 

689 The keys being held while scrolling.

690 

691 - `class Type:`

692 

693 An action to type in text.

694 

695 - `String text`

696 

697 The text to type.

698 

699 - `JsonValue; type "type"constant`

700 

701 Specifies the event type. For a type action, this property is always set to `type`.

702 

703 - `TYPE("type")`

704 

705 - `JsonValue;`

706 

707 - `JsonValue; type "wait"constant`

708 

709 Specifies the event type. For a wait action, this property is always set to `wait`.

710 

711 - `WAIT("wait")`

712 

713 - `Optional<List<ComputerAction>> actions`

714 

715 Flattened batched actions for `computer_use`. Each action includes an

716 `type` discriminator and action-specific fields.

717 

718 - `Click`

719 

720 - `Button button`

721 

722 Indicates which mouse button was pressed during the click. One of `left`, `right`, `wheel`, `back`, or `forward`.

723 

724 - `LEFT("left")`

725 

726 - `RIGHT("right")`

727 

728 - `WHEEL("wheel")`

729 

730 - `BACK("back")`

731 

732 - `FORWARD("forward")`

733 

734 - `JsonValue; type "click"constant`

735 

736 Specifies the event type. For a click action, this property is always `click`.

737 

738 - `CLICK("click")`

739 

740 - `long x`

741 

742 The x-coordinate where the click occurred.

743 

744 - `long y`

745 

746 The y-coordinate where the click occurred.

747 

748 - `Optional<List<String>> keys`

749 

750 The keys being held while clicking.

751 

752 - `DoubleClick`

753 

754 - `Optional<List<String>> keys`

755 

756 The keys being held while double-clicking.

757 

758 - `JsonValue; type "double_click"constant`

759 

760 Specifies the event type. For a double click action, this property is always set to `double_click`.

761 

762 - `DOUBLE_CLICK("double_click")`

763 

764 - `long x`

765 

766 The x-coordinate where the double click occurred.

767 

768 - `long y`

769 

770 The y-coordinate where the double click occurred.

771 

772 - `Drag`

773 

774 - `List<Path> path`

775 

776 An array of coordinates representing the path of the drag action. Coordinates will appear as an array of objects, eg

777 

778 ```

779 [

780 { x: 100, y: 200 },

781 { x: 200, y: 300 }

782 ]

783 ```

784 

785 - `long x`

786 

787 The x-coordinate.

788 

789 - `long y`

790 

791 The y-coordinate.

792 

793 - `JsonValue; type "drag"constant`

794 

795 Specifies the event type. For a drag action, this property is always set to `drag`.

796 

797 - `DRAG("drag")`

798 

799 - `Optional<List<String>> keys`

800 

801 The keys being held while dragging the mouse.

802 

803 - `Keypress`

804 

805 - `List<String> keys`

806 

807 The combination of keys the model is requesting to be pressed. This is an array of strings, each representing a key.

808 

809 - `JsonValue; type "keypress"constant`

810 

811 Specifies the event type. For a keypress action, this property is always set to `keypress`.

812 

813 - `KEYPRESS("keypress")`

814 

815 - `Move`

816 

817 - `JsonValue; type "move"constant`

818 

819 Specifies the event type. For a move action, this property is always set to `move`.

820 

821 - `MOVE("move")`

822 

823 - `long x`

824 

825 The x-coordinate to move to.

826 

827 - `long y`

828 

829 The y-coordinate to move to.

830 

831 - `Optional<List<String>> keys`

832 

833 The keys being held while moving the mouse.

834 

835 - `JsonValue;`

836 

837 - `JsonValue; type "screenshot"constant`

838 

839 Specifies the event type. For a screenshot action, this property is always set to `screenshot`.

840 

841 - `SCREENSHOT("screenshot")`

842 

843 - `Scroll`

844 

845 - `long scrollX`

846 

847 The horizontal scroll distance.

848 

849 - `long scrollY`

850 

851 The vertical scroll distance.

852 

853 - `JsonValue; type "scroll"constant`

854 

855 Specifies the event type. For a scroll action, this property is always set to `scroll`.

856 

857 - `SCROLL("scroll")`

858 

859 - `long x`

860 

861 The x-coordinate where the scroll occurred.

862 

863 - `long y`

864 

865 The y-coordinate where the scroll occurred.

866 

867 - `Optional<List<String>> keys`

868 

869 The keys being held while scrolling.

870 

871 - `Type`

872 

873 - `String text`

874 

875 The text to type.

876 

877 - `JsonValue; type "type"constant`

878 

879 Specifies the event type. For a type action, this property is always set to `type`.

880 

881 - `TYPE("type")`

882 

883 - `JsonValue;`

884 

885 - `JsonValue; type "wait"constant`

886 

887 Specifies the event type. For a wait action, this property is always set to `wait`.

888 

889 - `WAIT("wait")`

890 

891 - `ComputerCallOutput`

892 

893 - `String callId`

894 

895 The ID of the computer tool call that produced the output.

896 

897 - `ResponseComputerToolCallOutputScreenshot output`

898 

899 A computer screenshot image used with the computer use tool.

900 

901 - `JsonValue; type "computer_screenshot"constant`

902 

903 Specifies the event type. For a computer screenshot, this property is

904 always set to `computer_screenshot`.

905 

906 - `COMPUTER_SCREENSHOT("computer_screenshot")`

907 

908 - `Optional<String> fileId`

909 

910 The identifier of an uploaded file that contains the screenshot.

911 

912 - `Optional<String> imageUrl`

913 

914 The URL of the screenshot image.

915 

916 - `JsonValue; type "computer_call_output"constant`

917 

918 The type of the computer tool call output. Always `computer_call_output`.

919 

920 - `COMPUTER_CALL_OUTPUT("computer_call_output")`

921 

922 - `Optional<String> id`

923 

924 The ID of the computer tool call output.

925 

926 - `Optional<List<AcknowledgedSafetyCheck>> acknowledgedSafetyChecks`

927 

928 The safety checks reported by the API that have been acknowledged by the developer.

929 

930 - `String id`

931 

932 The ID of the pending safety check.

933 

934 - `Optional<String> code`

935 

936 The type of the pending safety check.

937 

938 - `Optional<String> message`

939 

940 Details about the pending safety check.

941 

942 - `Optional<Status> status`

943 

944 The status of the message input. One of `in_progress`, `completed`, or `incomplete`. Populated when input items are returned via API.

945 

946 - `IN_PROGRESS("in_progress")`

947 

948 - `COMPLETED("completed")`

949 

950 - `INCOMPLETE("incomplete")`

951 

952 - `class ResponseFunctionWebSearch:`

953 

954 The results of a web search tool call. See the

955 [web search guide](https://platform.openai.com/docs/guides/tools-web-search) for more information.

956 

957 - `String id`

958 

959 The unique ID of the web search tool call.

960 

961 - `Action action`

962 

963 An object describing the specific action taken in this web search call.

964 Includes details on how the model used the web (search, open_page, find_in_page).

965 

966 - `class Search:`

967 

968 Action type "search" - Performs a web search query.

969 

970 - `JsonValue; type "search"constant`

971 

972 The action type.

973 

974 - `SEARCH("search")`

975 

976 - `Optional<List<String>> queries`

977 

978 The search queries.

979 

980 - `Optional<String> query`

981 

982 The search query.

983 

984 - `Optional<List<Source>> sources`

985 

986 The sources used in the search.

987 

988 - `JsonValue; type "url"constant`

989 

990 The type of source. Always `url`.

991 

992 - `URL("url")`

993 

994 - `String url`

995 

996 The URL of the source.

997 

998 - `class OpenPage:`

999 

1000 Action type "open_page" - Opens a specific URL from search results.

1001 

1002 - `JsonValue; type "open_page"constant`

1003 

1004 The action type.

1005 

1006 - `OPEN_PAGE("open_page")`

1007 

1008 - `Optional<String> url`

1009 

1010 The URL opened by the model.

1011 

1012 - `class FindInPage:`

1013 

1014 Action type "find_in_page": Searches for a pattern within a loaded page.

1015 

1016 - `String pattern`

1017 

1018 The pattern or text to search for within the page.

1019 

1020 - `JsonValue; type "find_in_page"constant`

1021 

1022 The action type.

1023 

1024 - `FIND_IN_PAGE("find_in_page")`

1025 

1026 - `String url`

1027 

1028 The URL of the page searched for the pattern.

1029 

1030 - `Status status`

1031 

1032 The status of the web search tool call.

1033 

1034 - `IN_PROGRESS("in_progress")`

1035 

1036 - `SEARCHING("searching")`

1037 

1038 - `COMPLETED("completed")`

1039 

1040 - `FAILED("failed")`

1041 

1042 - `JsonValue; type "web_search_call"constant`

1043 

1044 The type of the web search tool call. Always `web_search_call`.

1045 

1046 - `WEB_SEARCH_CALL("web_search_call")`

1047 

1048 - `class ResponseFunctionToolCall:`

1049 

1050 A tool call to run a function. See the

1051 [function calling guide](https://platform.openai.com/docs/guides/function-calling) for more information.

1052 

1053 - `String arguments`

1054 

1055 A JSON string of the arguments to pass to the function.

1056 

1057 - `String callId`

1058 

1059 The unique ID of the function tool call generated by the model.

1060 

1061 - `String name`

1062 

1063 The name of the function to run.

1064 

1065 - `JsonValue; type "function_call"constant`

1066 

1067 The type of the function tool call. Always `function_call`.

1068 

1069 - `FUNCTION_CALL("function_call")`

1070 

1071 - `Optional<String> id`

1072 

1073 The unique ID of the function tool call.

1074 

1075 - `Optional<Caller> caller`

1076 

1077 The execution context that produced this tool call.

1078 

1079 - `JsonValue;`

1080 

1081 - `JsonValue; type "direct"constant`

1082 

1083 - `DIRECT("direct")`

1084 

1085 - `class Program:`

1086 

1087 - `String callerId`

1088 

1089 The call ID of the program item that produced this tool call.

1090 

1091 - `JsonValue; type "program"constant`

1092 

1093 - `PROGRAM("program")`

1094 

1095 - `Optional<String> namespace`

1096 

1097 The namespace of the function to run.

1098 

1099 - `Optional<Status> status`

1100 

1101 The status of the item. One of `in_progress`, `completed`, or

1102 `incomplete`. Populated when items are returned via API.

1103 

1104 - `IN_PROGRESS("in_progress")`

1105 

1106 - `COMPLETED("completed")`

1107 

1108 - `INCOMPLETE("incomplete")`

1109 

1110 - `FunctionCallOutput`

1111 

1112 - `String callId`

1113 

1114 The unique ID of the function tool call generated by the model.

1115 

1116 - `Output output`

1117 

1118 Text, image, or file output of the function tool call.

1119 

1120 - `String`

1121 

1122 - `List<ResponseFunctionCallOutputItem>`

1123 

1124 - `class ResponseInputTextContent:`

1125 

1126 A text input to the model.

1127 

1128 - `String text`

1129 

1130 The text input to the model.

1131 

1132 - `JsonValue; type "input_text"constant`

1133 

1134 The type of the input item. Always `input_text`.

1135 

1136 - `INPUT_TEXT("input_text")`

1137 

1138 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

1139 

1140 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

1141 

1142 - `JsonValue; mode "explicit"constant`

1143 

1144 The breakpoint mode. Always `explicit`.

1145 

1146 - `EXPLICIT("explicit")`

1147 

1148 - `class ResponseInputImageContent:`

1149 

1150 An image input to the model. Learn about [image inputs](https://platform.openai.com/docs/guides/vision)

1151 

1152 - `JsonValue; type "input_image"constant`

1153 

1154 The type of the input item. Always `input_image`.

1155 

1156 - `INPUT_IMAGE("input_image")`

1157 

1158 - `Optional<Detail> detail`

1159 

1160 The detail level of the image to be sent to the model. One of `high`, `low`, `auto`, or `original`. Defaults to `auto`.

1161 

1162 - `LOW("low")`

1163 

1164 - `HIGH("high")`

1165 

1166 - `AUTO("auto")`

1167 

1168 - `ORIGINAL("original")`

1169 

1170 - `Optional<String> fileId`

1171 

1172 The ID of the file to be sent to the model.

1173 

1174 - `Optional<String> imageUrl`

1175 

1176 The URL of the image to be sent to the model. A fully qualified URL or base64 encoded image in a data URL.

1177 

1178 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

1179 

1180 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

1181 

1182 - `JsonValue; mode "explicit"constant`

1183 

1184 The breakpoint mode. Always `explicit`.

1185 

1186 - `EXPLICIT("explicit")`

1187 

1188 - `class ResponseInputFileContent:`

1189 

1190 A file input to the model.

1191 

1192 - `JsonValue; type "input_file"constant`

1193 

1194 The type of the input item. Always `input_file`.

1195 

1196 - `INPUT_FILE("input_file")`

1197 

1198 - `Optional<Detail> detail`

1199 

1200 The detail level of the file to be sent to the model. Use `auto` to let the system select the detail level; for GPT-5.6 and later models, `auto` uses high-quality rendering, which may increase input token usage. Use `low` for lower-cost rendering, or `high` to render the file at higher quality. Defaults to `auto`.

1201 

1202 - `AUTO("auto")`

1203 

1204 - `LOW("low")`

1205 

1206 - `HIGH("high")`

1207 

1208 - `Optional<String> fileData`

1209 

1210 The base64-encoded data of the file to be sent to the model.

1211 

1212 - `Optional<String> fileId`

1213 

1214 The ID of the file to be sent to the model.

1215 

1216 - `Optional<String> fileUrl`

1217 

1218 The URL of the file to be sent to the model.

1219 

1220 - `Optional<String> filename`

1221 

1222 The name of the file to be sent to the model.

1223 

1224 - `Optional<PromptCacheBreakpoint> promptCacheBreakpoint`

1225 

1226 Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's `prompt_cache_options.ttl`; the boundary is not rounded to a token block.

1227 

1228 - `JsonValue; mode "explicit"constant`

1229 

1230 The breakpoint mode. Always `explicit`.

1231 

1232 - `EXPLICIT("explicit")`

1233 

1234 - `JsonValue; type "function_call_output"constant`

1235 

1236 The type of the function tool call output. Always `function_call_output`.

1237 

1238 - `FUNCTION_CALL_OUTPUT("function_call_output")`

1239 

1240 - `Optional<String> id`

1241 

1242 The unique ID of the function tool call output. Populated when this item is returned via API.

1243 

1244 - `Optional<Caller> caller`

1245 

1246 The execution context that produced this tool call.

1247 

1248 - `JsonValue;`

1249 

1250 - `JsonValue; type "direct"constant`

1251 

1252 The caller type. Always `direct`.

1253 

1254 - `DIRECT("direct")`

1255 

1256 - `class Program:`

1257 

1258 - `String callerId`

1259 

1260 The call ID of the program item that produced this tool call.

1261 

1262 - `JsonValue; type "program"constant`

1263 

1264 The caller type. Always `program`.

1265 

1266 - `PROGRAM("program")`

1267 

1268 - `Optional<Status> status`

1269 

1270 The status of the item. One of `in_progress`, `completed`, or `incomplete`. Populated when items are returned via API.

1271 

1272 - `IN_PROGRESS("in_progress")`

1273 

1274 - `COMPLETED("completed")`

1275 

1276 - `INCOMPLETE("incomplete")`

1277 

1278 - `ToolSearchCall`

1279 

1280 - `JsonValue arguments`

1281 

1282 The arguments supplied to the tool search call.

1283 

1284 - `JsonValue; type "tool_search_call"constant`

1285 

1286 The item type. Always `tool_search_call`.

1287 

1288 - `TOOL_SEARCH_CALL("tool_search_call")`

1289 

1290 - `Optional<String> id`

1291 

1292 The unique ID of this tool search call.

1293 

1294 - `Optional<String> callId`

1295 

1296 The unique ID of the tool search call generated by the model.

1297 

1298 - `Optional<Execution> execution`

1299 

1300 Whether tool search was executed by the server or by the client.

1301 

1302 - `SERVER("server")`

1303 

1304 - `CLIENT("client")`

1305 

1306 - `Optional<Status> status`

1307 

1308 The status of the tool search call.

1309 

1310 - `IN_PROGRESS("in_progress")`

1311 

1312 - `COMPLETED("completed")`

1313 

1314 - `INCOMPLETE("incomplete")`

1315 

1316 - `class ResponseToolSearchOutputItemParam:`

1317 

1318 - `List<Tool> tools`

1319 

1320 The loaded tool definitions returned by the tool search output.

1321 

1322 - `class FunctionTool:`

1323 

1324 Defines a function in your own code the model can choose to call. Learn more about [function calling](https://platform.openai.com/docs/guides/function-calling).

1325 

1326 - `String name`

1327 

1328 The name of the function to call.

1329 

1330 - `Optional<Parameters> parameters`

1331 

1332 A JSON schema object describing the parameters of the function.

1333 

1334 - `Optional<Boolean> strict`

1335 

1336 Whether strict parameter validation is enforced for this function tool.

1337 

1338 - `JsonValue; type "function"constant`

1339 

1340 The type of the function tool. Always `function`.

1341 

1342 - `FUNCTION("function")`

1343 

1344 - `Optional<List<AllowedCaller>> allowedCallers`

1345 

1346 The tool invocation context(s).

1347 

1348 - `DIRECT("direct")`

1349 

1350 - `PROGRAMMATIC("programmatic")`

1351 

1352 - `Optional<Boolean> deferLoading`

1353 

1354 Whether this function is deferred and loaded via tool search.

1355 

1356 - `Optional<String> description`

1357 

1358 A description of the function. Used by the model to determine whether or not to call the function.

1359 

1360 - `Optional<OutputSchema> outputSchema`

1361 

1362 A JSON schema object describing the JSON value encoded in string outputs for this function.

1363 

1364 - `class FileSearchTool:`

1365 

1366 A tool that searches for relevant content from uploaded files. Learn more about the [file search tool](https://platform.openai.com/docs/guides/tools-file-search).

1367 

1368 - `JsonValue; type "file_search"constant`

1369 

1370 The type of the file search tool. Always `file_search`.

1371 

1372 - `FILE_SEARCH("file_search")`

1373 

1374 - `List<String> vectorStoreIds`

1375 

1376 The IDs of the vector stores to search.

1377 

1378 - `Optional<Filters> filters`

1379 

1380 A filter to apply.

1381 

1382 - `class ComparisonFilter:`

1383 

1384 A filter used to compare a specified attribute key to a given value using a defined comparison operation.

1385 

1386 - `String key`

1387 

1388 The key to compare against the value.

1389 

1390 - `Type type`

1391 

1392 Specifies the comparison operator: `eq`, `ne`, `gt`, `gte`, `lt`, `lte`, `in`, `nin`.

1393 

1394 - `eq`: equals

1395 - `ne`: not equal

1396 - `gt`: greater than

1397 - `gte`: greater than or equal

1398 - `lt`: less than

1399 - `lte`: less than or equal

1400 - `in`: in

1401 - `nin`: not in

1402 

1403 - `EQ("eq")`

1404 

1405 - `NE("ne")`

1406 

1407 - `GT("gt")`

1408 

1409 - `GTE("gte")`

1410 

1411 - `LT("lt")`

1412 

1413 - `LTE("lte")`

1414 

1415 - `IN("in")`

1416 

1417 - `NIN("nin")`

1418 

1419 - `Value value`

1420 

1421 The value to compare against the attribute key; supports string, number, or boolean types.

1422 

1423 - `String`

1424 

1425 - `double`

1426 

1427 - `boolean`

1428 

1429 - `List<ComparisonFilterValueItem>`

1430 

1431 - `String`

1432 

1433 - `double`

1434 

1435 - `class CompoundFilter:`

1436 

1437 Combine multiple filters using `and` or `or`.

1438 

1439 - `List<Filter> filters`

1440 

1441 Array of filters to combine. Items can be `ComparisonFilter` or `CompoundFilter`.

1442 

1443 - `class ComparisonFilter:`

1444 

1445 A filter used to compare a specified attribute key to a given value using a defined comparison operation.

1446 

1447 - `JsonValue`

1448 

1449 - `Type type`

1450 

1451 Type of operation: `and` or `or`.

1452 

1453 - `AND("and")`

1454 

1455 - `OR("or")`

1456 

1457 - `Optional<Long> maxNumResults`

1458 

1459 The maximum number of results to return. This number should be between 1 and 50 inclusive.

1460 

1461 - `Optional<RankingOptions> rankingOptions`

1462 

1463 Ranking options for search.

1464 

1465 - `Optional<HybridSearch> hybridSearch`

1466 

1467 Weights that control how reciprocal rank fusion balances semantic embedding matches versus sparse keyword matches when hybrid search is enabled.

1468 

1469 - `double embeddingWeight`

1470 

1471 The weight of the embedding in the reciprocal ranking fusion.

1472 

1473 - `double textWeight`

1474 

1475 The weight of the text in the reciprocal ranking fusion.

1476 

1477 - `Optional<Ranker> ranker`

1478 

1479 The ranker to use for the file search.

1480 

1481 - `AUTO("auto")`

1482 

1483 - `DEFAULT_2024_11_15("default-2024-11-15")`

1484 

1485 - `Optional<Double> scoreThreshold`

1486 

1487 The score threshold for the file search, a number between 0 and 1. Numbers closer to 1 will attempt to return only the most relevant results, but may return fewer results.

1488 

1489 - `class ComputerTool:`

1490 

1491 A tool that controls a virtual computer. Learn more about the [computer tool](https://platform.openai.com/docs/guides/tools-computer-use).

1492 

1493 - `JsonValue; type "computer"constant`

1494 

1495 The type of the computer tool. Always `computer`.

1496 

1497 - `COMPUTER("computer")`

1498 

1499 - `class ComputerUsePreviewTool:`

1500 

1501 A tool that controls a virtual computer. Learn more about the [computer tool](https://platform.openai.com/docs/guides/tools-computer-use).

1502 

1503 - `long displayHeight`

1504 

1505 The height of the computer display.

1506 

1507 - `long displayWidth`

1508 

1509 The width of the computer display.

1510 

1511 - `Environment environment`

1512 

1513 The type of computer environment to control.

1514 

1515 - `WINDOWS("windows")`

1516 

1517 - `MAC("mac")`

1518 

1519 - `LINUX("linux")`

1520 

1521 - `UBUNTU("ubuntu")`

1522 

1523 - `BROWSER("browser")`

1524 

1525 - `JsonValue; type "computer_use_preview"constant`

1526 

1527 The type of the computer use tool. Always `computer_use_preview`.

1528 

1529 - `COMPUTER_USE_PREVIEW("computer_use_preview")`

1530 

1531 - `class WebSearchTool:`

1532 

1533 Search the Internet for sources related to the prompt. Learn more about the

1534 [web search tool](https://platform.openai.com/docs/guides/tools-web-search).

1535 

1536 - `Type type`

1537 

1538 The type of the web search tool. One of `web_search` or `web_search_2025_08_26`.

1539 

1540 - `WEB_SEARCH("web_search")`

1541 

1542 - `WEB_SEARCH_2025_08_26("web_search_2025_08_26")`

1543 

1544 - `Optional<Filters> filters`

1545 

1546 Filters for the search.

1547 

1548 - `Optional<List<String>> allowedDomains`

1549 

1550 Allowed domains for the search. If not provided, all domains are allowed.

1551 Subdomains of the provided domains are allowed as well.

1552 

1553 Example: `["pubmed.ncbi.nlm.nih.gov"]`

1554 

1555 - `Optional<SearchContextSize> searchContextSize`

1556 

1557 High level guidance for the amount of context window space to use for the search. One of `low`, `medium`, or `high`. `medium` is the default.

1558 

1559 - `LOW("low")`

1560 

1561 - `MEDIUM("medium")`

1562 

1563 - `HIGH("high")`

1564 

1565 - `Optional<UserLocation> userLocation`

1566 

1567 The approximate location of the user.

1568 

1569 - `Optional<String> city`

1570 

1571 Free text input for the city of the user, e.g. `San Francisco`.

1572 

1573 - `Optional<String> country`

1574 

1575 The two-letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1) of the user, e.g. `US`.

1576 

1577 - `Optional<String> region`

1578 

1579 Free text input for the region of the user, e.g. `California`.

1580 

1581 - `Optional<String> timezone`

1582 

1583 The [IANA timezone](https://timeapi.io/documentation/iana-timezones) of the user, e.g. `America/Los_Angeles`.

1584 

1585 - `Optional<Type> type`

1586 

1587 The type of location approximation. Always `approximate`.

1588 

1589 - `APPROXIMATE("approximate")`

1590 

1591 - `Mcp`

1592 

1593 - `String serverLabel`

1594 

1595 A label for this MCP server, used to identify it in tool calls.

1596 

1597 - `JsonValue; type "mcp"constant`

1598 

1599 The type of the MCP tool. Always `mcp`.

1600 

1601 - `MCP("mcp")`

1602 

1603 - `Optional<List<AllowedCaller>> allowedCallers`

1604 

1605 The tool invocation context(s).

1606 

1607 - `DIRECT("direct")`

1608 

1609 - `PROGRAMMATIC("programmatic")`

1610 

1611 - `Optional<AllowedTools> allowedTools`

1612 

1613 List of allowed tool names or a filter object.

1614 

1615 - `List<String>`

1616 

1617 - `class McpToolFilter:`

1618 

1619 A filter object to specify which tools are allowed.

1620 

1621 - `Optional<Boolean> readOnly`

1622 

1623 Indicates whether or not a tool modifies data or is read-only. If an

1624 MCP server is [annotated with `readOnlyHint`](https://modelcontextprotocol.io/specification/2025-06-18/schema#toolannotations-readonlyhint),

1625 it will match this filter.

1626 

1627 - `Optional<List<String>> toolNames`

1628 

1629 List of allowed tool names.

1630 

1631 - `Optional<String> authorization`

1632 

1633 An OAuth access token that can be used with a remote MCP server, either

1634 with a custom MCP server URL or a service connector. Your application

1635 must handle the OAuth authorization flow and provide the token here.

1636 

1637 - `Optional<ConnectorId> connectorId`

1638 

1639 Identifier for service connectors, like those available in ChatGPT. One of

1640 `server_url`, `connector_id`, or `tunnel_id` must be provided. Learn more

1641 about service connectors [here](https://platform.openai.com/docs/guides/tools-remote-mcp#connectors).

1642 

1643 Currently supported `connector_id` values are:

1644 

1645 - Dropbox: `connector_dropbox`

1646 - Gmail: `connector_gmail`

1647 - Google Calendar: `connector_googlecalendar`

1648 - Google Drive: `connector_googledrive`

1649 - Microsoft Teams: `connector_microsoftteams`

1650 - Outlook Calendar: `connector_outlookcalendar`

1651 - Outlook Email: `connector_outlookemail`

1652 - SharePoint: `connector_sharepoint`

1653 

1654 - `CONNECTOR_DROPBOX("connector_dropbox")`

1655 

1656 - `CONNECTOR_GMAIL("connector_gmail")`

1657 

1658 - `CONNECTOR_GOOGLECALENDAR("connector_googlecalendar")`

1659 

1660 - `CONNECTOR_GOOGLEDRIVE("connector_googledrive")`

1661 

1662 - `CONNECTOR_MICROSOFTTEAMS("connector_microsoftteams")`

1663 

1664 - `CONNECTOR_OUTLOOKCALENDAR("connector_outlookcalendar")`

1665 

1666 - `CONNECTOR_OUTLOOKEMAIL("connector_outlookemail")`

1667 

1668 - `CONNECTOR_SHAREPOINT("connector_sharepoint")`

1669 

1670 - `Optional<Boolean> deferLoading`

1671 

1672 Whether this MCP tool is deferred and discovered via tool search.

1673 

1674 - `Optional<Headers> headers`

1675 

1676 Optional HTTP headers to send to the MCP server. Use for authentication

1677 or other purposes.

1678 

1679 - `Optional<RequireApproval> requireApproval`

1680 

1681 Specify which of the MCP server's tools require approval.

1682 

1683 - `class McpToolApprovalFilter:`

1684 

1685 Specify which of the MCP server's tools require approval. Can be

1686 `always`, `never`, or a filter object associated with tools

1687 that require approval.

1688 

1689 - `Optional<Always> always`

1690 

1691 A filter object to specify which tools are allowed.

1692 

1693 - `Optional<Boolean> readOnly`

1694 

1695 Indicates whether or not a tool modifies data or is read-only. If an

1696 MCP server is [annotated with `readOnlyHint`](https://modelcontextprotocol.io/specification/2025-06-18/schema#toolannotations-readonlyhint),

1697 it will match this filter.

1698 

1699 - `Optional<List<String>> toolNames`

1700 

1701 List of allowed tool names.

1702 

1703 - `Optional<Never> never`

1704 

1705 A filter object to specify which tools are allowed.

1706 

1707 - `Optional<Boolean> readOnly`

1708 

1709 Indicates whether or not a tool modifies data or is read-only. If an

1710 MCP server is [annotated with `readOnlyHint`](https://modelcontextprotocol.io/specification/2025-06-18/schema#toolannotations-readonlyhint),

1711 it will match this filter.

1712 

1713 - `Optional<List<String>> toolNames`

1714 

1715 List of allowed tool names.

1716 

1717 - `enum McpToolApprovalSetting:`

1718 

1719 Specify a single approval policy for all tools. One of `always` or

1720 `never`. When set to `always`, all tools will require approval. When

1721 set to `never`, all tools will not require approval.

1722 

1723 - `ALWAYS("always")`

1724 

1725 - `NEVER("never")`

1726 

1727 - `Optional<String> serverDescription`

1728 

1729 Optional description of the MCP server, used to provide more context.

1730 

1731 - `Optional<String> serverUrl`

1732 

1733 The URL for the MCP server. One of `server_url`, `connector_id`, or

1734 `tunnel_id` must be provided.

1735 

1736 - `Optional<String> tunnelId`

1737 

1738 The Secure MCP Tunnel ID to use instead of a direct server URL. One of

1739 `server_url`, `connector_id`, or `tunnel_id` must be provided.

1740 

1741 - `CodeInterpreter`

1742 

1743 - `Container container`

1744 

1745 The code interpreter container. Can be a container ID or an object that

1746 specifies uploaded file IDs to make available to your code, along with an

1747 optional `memory_limit` setting.

1748 

1749 - `String`

1750 

1751 - `class CodeInterpreterToolAuto:`

1752 

1753 Configuration for a code interpreter container. Optionally specify the IDs of the files to run the code on.

1754 

1755 - `JsonValue; type "auto"constant`

1756 

1757 Always `auto`.

1758 

1759 - `AUTO("auto")`

1760 

1761 - `Optional<List<String>> fileIds`

1762 

1763 An optional list of uploaded files to make available to your code.

1764 

1765 - `Optional<MemoryLimit> memoryLimit`

1766 

1767 The memory limit for the code interpreter container.

1768 

1769 - `_1G("1g")`

1770 

1771 - `_4G("4g")`

1772 

1773 - `_16G("16g")`

1774 

1775 - `_64G("64g")`

1776 

1777 - `Optional<NetworkPolicy> networkPolicy`

1778 

1779 Network access policy for the container.

1780 

1781 - `class ContainerNetworkPolicyDisabled:`

1782 

1783 - `JsonValue; type "disabled"constant`

1784 

1785 Disable outbound network access. Always `disabled`.

1786 

1787 - `DISABLED("disabled")`

1788 

1789 - `class ContainerNetworkPolicyAllowlist:`

1790 

1791 - `List<String> allowedDomains`

1792 

1793 A list of allowed domains when type is `allowlist`.

1794 

1795 - `JsonValue; type "allowlist"constant`

1796 

1797 Allow outbound network access only to specified domains. Always `allowlist`.

1798 

1799 - `ALLOWLIST("allowlist")`

1800 

1801 - `Optional<List<ContainerNetworkPolicyDomainSecret>> domainSecrets`

1802 

1803 Optional domain-scoped secrets for allowlisted domains.

1804 

1805 - `String domain`

1806 

1807 The domain associated with the secret.

1808 

1809 - `String name`

1810 

1811 The name of the secret to inject for the domain.

1812 

1813 - `String value`

1814 

1815 The secret value to inject for the domain.

1816 

1817 - `JsonValue; type "code_interpreter"constant`

1818 

1819 The type of the code interpreter tool. Always `code_interpreter`.

1820 

1821 - `CODE_INTERPRETER("code_interpreter")`

1822 

1823 - `Optional<List<AllowedCaller>> allowedCallers`

1824 

1825 The tool invocation context(s).

1826 

1827 - `DIRECT("direct")`

1828 

1829 - `PROGRAMMATIC("programmatic")`

1830 

1831 - `JsonValue;`

1832 

1833 - `JsonValue; type "programmatic_tool_calling"constant`

1834 

1835 The type of the tool. Always `programmatic_tool_calling`.

1836 

1837 - `PROGRAMMATIC_TOOL_CALLING("programmatic_tool_calling")`

1838 

1839 - `ImageGeneration`

1840 

1841 - `JsonValue; type "image_generation"constant`

1842 

1843 The type of the image generation tool. Always `image_generation`.

1844 

1845 - `IMAGE_GENERATION("image_generation")`

1846 

1847 - `Optional<Action> action`

1848 

1849 Whether to generate a new image or edit an existing image. Default: `auto`.

1850 

1851 - `GENERATE("generate")`

1852 

1853 - `EDIT("edit")`

1854 

1855 - `AUTO("auto")`

1856 

1857 - `Optional<Background> background`

1858 

1859 Allows to set transparency for the background of the generated image(s).

1860 This parameter is only supported for GPT image models that support

1861 transparent backgrounds. Must be one of `transparent`, `opaque`, or

1862 `auto` (default value). When `auto` is used, the model will

1863 automatically determine the best background for the image.

1864 

1865 `gpt-image-2` and `gpt-image-2-2026-04-21` do not support

1866 transparent backgrounds. Requests with `background` set to

1867 `transparent` will return an error for these models; use `opaque` or

1868 `auto` instead.

1869 

1870 If `transparent`, the output format needs to support transparency,

1871 so it should be set to either `png` (default value) or `webp`.

1872 

1873 - `TRANSPARENT("transparent")`

1874 

1875 - `OPAQUE("opaque")`

1876 

1877 - `AUTO("auto")`

1878 

1879 - `Optional<InputFidelity> inputFidelity`

1880 

1881 Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for `gpt-image-1` and `gpt-image-1.5` and later models, unsupported for `gpt-image-1-mini`. Supports `high` and `low`. Defaults to `low`.

1882 

1883 - `HIGH("high")`

1884 

1885 - `LOW("low")`

1886 

1887 - `Optional<InputImageMask> inputImageMask`

1888 

1889 Optional mask for inpainting. Contains `image_url`

1890 (string, optional) and `file_id` (string, optional).

1891 

1892 - `Optional<String> fileId`

1893 

1894 File ID for the mask image.

1895 

1896 - `Optional<String> imageUrl`

1897 

1898 Base64-encoded mask image.

1899 

1900 - `Optional<Model> model`

1901 

1902 The image generation model to use. Default: `gpt-image-1`.

1903 

1904 - `GPT_IMAGE_1("gpt-image-1")`

1905 

1906 - `GPT_IMAGE_1_MINI("gpt-image-1-mini")`

1907 

1908 - `GPT_IMAGE_2("gpt-image-2")`

1909 

1910 - `GPT_IMAGE_2_2026_04_21("gpt-image-2-2026-04-21")`

1911 

1912 - `GPT_IMAGE_1_5("gpt-image-1.5")`

1913 

1914 - `CHATGPT_IMAGE_LATEST("chatgpt-image-latest")`

1915 

1916 - `Optional<Moderation> moderation`

1917 

1918 Moderation level for the generated image. Default: `auto`.

1919 

1920 - `AUTO("auto")`

1921 

1922 - `LOW("low")`

1923 

1924 - `Optional<Long> outputCompression`

1925 

1926 Compression level for the output image. Default: 100.

1927 

1928 - `Optional<OutputFormat> outputFormat`

1929 

1930 The output format of the generated image. One of `png`, `webp`, or

1931 `jpeg`. Default: `png`.

1932 

1933 - `PNG("png")`

1934 

1935 - `WEBP("webp")`

1936 

1937 - `JPEG("jpeg")`

1938 

1939 - `Optional<Long> partialImages`

1940 

1941 Number of partial images to generate in streaming mode, from 0 (default value) to 3.

1942 

1943 - `Optional<Quality> quality`

1944 

1945 The quality of the generated image. One of `low`, `medium`, `high`,

1946 or `auto`. Default: `auto`.

1947 

1948 - `LOW("low")`

1949 

1950 - `MEDIUM("medium")`

1951 

1952 - `HIGH("high")`

1953 

1954 - `AUTO("auto")`

1955 

1956 - `Optional<Size> size`

1957 

1958 The size of the generated images. For `gpt-image-2` and `gpt-image-2-2026-04-21`, arbitrary resolutions are supported as `WIDTHxHEIGHT` strings, for example `1536x864`. Width and height must both be divisible by 16 and the requested aspect ratio must be between 1:3 and 3:1. Resolutions above `2560x1440` are experimental, and the maximum supported resolution is `3840x2160`. The requested size must also satisfy the model's current pixel and edge limits. The standard sizes `1024x1024`, `1536x1024`, and `1024x1536` are supported by the GPT image models; `auto` is supported for models that allow automatic sizing. For `dall-e-2`, use one of `256x256`, `512x512`, or `1024x1024`. For `dall-e-3`, use one of `1024x1024`, `1792x1024`, or `1024x1792`.

1959 

1960 - `_1024X1024("1024x1024")`

1961 

1962 - `_1024X1536("1024x1536")`

1963 

1964 - `_1536X1024("1536x1024")`

1965 

1966 - `AUTO("auto")`

1967 

1968 - `JsonValue;`

1969 

1970 - `JsonValue; type "local_shell"constant`

1971 

1972 The type of the local shell tool. Always `local_shell`.

1973 

1974 - `LOCAL_SHELL("local_shell")`

1975 

1976 - `class FunctionShellTool:`

1977 

1978 A tool that allows the model to execute shell commands.

1979 

1980 - `JsonValue; type "shell"constant`

1981 

1982 The type of the shell tool. Always `shell`.

1983 

1984 - `SHELL("shell")`

1985 

1986 - `Optional<List<AllowedCaller>> allowedCallers`

1987 

1988 The tool invocation context(s).

1989 

1990 - `DIRECT("direct")`

1991 

1992 - `PROGRAMMATIC("programmatic")`

1993 

1994 - `Optional<Environment> environment`

1995 

1996 - `class ContainerAuto:`