------------------------------------------------------------------- Sun Mar 8 20:34:17 UTC 2026 - Dirk Müller - update to 2.26.0: * **api:** The GA ComputerTool now uses the CompuerTool class. The 'computer_use_preview' tool is moved to ComputerUsePreview * **api:** gpt-5.4, tool search tool, and new computer tool * **api:** remove prompt_cache_key param from responses, phase field from message types * **api:** add phase * **api:** add gpt-realtime-1.5 and gpt-audio-1.5 model options to realtime calls * **api:** websockets for responses api * **internal:** add request options to SSE classes * update mock server docs * **api:** add batch size limit to file_batches parameter descriptions * **api:** enhance method descriptions across audio, chat, realtime, skills, uploads, videos * **api:** update safety_identifier documentation in chat completions and responses * **structured outputs:** resolve memory leak in parse methods * **webhooks:** preserve method visibility for compatibility checks * **internal:** fix lint error on Python 3.14 * **api:** support for images in batch api * **api:** skills and hosted shell * **internal:** bump dependencies * **api:** add context_management to responses * **api:** responses context_management * **api:** add shell_call_output status field * **api:** image generation actions for responses; ResponseFunctionCallArgumentsDoneEvent.name * **client:** add custom JSON encoder for extended type support * **client:** undo change to web search Find action * **client:** update type for `find_in_page` action * **api:** api update * **api:** api updates * **client:** add support for binary request streaming * **api:** mark assistants as deprecated * **ci:** upgrade `actions/github-script` * **internal:** update `actions/checkout` version ------------------------------------------------------------------- Fri Jan 9 21:52:32 UTC 2026 - Marius Grossu - Update to version 2.14.0: * api: slugs for new audio models; make all model params accept strings (e517792) * use async_to_httpx_files in patch method (a6af9ee) * internal: add --fix argument to lint script - complete changelog here: * https://github.com/openai/openai-python/compare/v2.8.1...v2.13.0 ------------------------------------------------------------------- Tue Nov 25 12:54:53 UTC 2025 - Marius Grossu - Update to version 2.8.1: * api: align types of input items / output items for typescript - Update to version 2.8.0: * api: gpt 5.1 * compat: update signatures of model_dump and model_dump_json for Pydantic v1 - Update to version 2.7.2: * compat with Python 3.14 - Update to version to 2.7.0: * client: close streams without requiring full consumption * readme: update realtime examples * uploads: avoid file handle leak * api: Realtime API token_limits, Hybrid searching ranking options * api: remove InputAudio from ResponseInputContent - complete changelog here: * https://github.com/openai/openai-python/compare/v1.91.0...v2.8.1 ------------------------------------------------------------------- Thu Jun 26 17:50:11 UTC 2025 - aline.werner@suse.com - Update to version 1.91.0: * release: 1.91.0 (#2423) * release: 1.90.0 (#2420) * release: 1.89.0 * feat(client): add support for aiohttp * fix(tests): fix: tests which call HTTP endpoints directly with the example parameters * chore(readme): update badges * release: 1.88.0 * chore(ci): enable for pull requests * chore(internal): minor formatting * feat(api): manual updates * release: 1.87.0 (#2410) * release: 1.86.0 (#2405) * release: 1.85.0 * feat(api): Add tools and structured outputs to evals * fix(responses): support raw responses for `parse()` * release: 1.84.0 (#2395) * release: 1.83.0 (#2393) * release: 1.82.1 * fix(responses): don't include `parsed_arguments` when re-serialising * chore(internal): fix release workflows * codegen metadata * release: 1.82.0 (#2372) * release: 1.81.0 (#2368) * release: 1.80.0 (#2367) * release: 1.79.0 * feat(api): further updates for evals API * fix: fix create audio transcription endpoint * feat(api): Updating Assistants and Evals API schemas * feat(api): manual updates * feat(api): responses x eval api * chore(ci): fix installation instructions * chore(ci): upload sdks to package manager * release: 1.78.1 * fix(internal): fix linting due to broken __test__ annotation * fix(package): support direct resource imports * release: 1.78.0 * feat(api): Add reinforcement fine-tuning api support * chore(internal): update proxy tests * fix: ignore errors in isinstance() calls on LazyProxy subclasses (#2343) * chore: use lazy imports for resources * chore: use lazy imports for module level client * release: 1.77.0 * feat(api): add image sizes, reasoning encryption * chore: only strip leading whitespace * fix(parsing): handle whitespace only strings (#2007) * release: 1.76.2 * chore(api): API spec cleanup * release: 1.76.1 * chore: broadly detect json family of content-type headers * chore(ci): only use depot for staging repos * chore(ci): run on more branches and use depot runners * release: 1.76.0 * feat(api): adding new image model support * chore(internal): minor formatting changes * fix(pydantic v1): more robust ModelField.annotation check * fix(pydantic v1): more robust `ModelField.annotation` check (#2163) * chore(internal): refactor retries to not use recursion * chore(internal): fix list file params * chore(internal): import reformatting * chore(ci): add timeout thresholds for CI jobs * chore: update completion parse signature * chore(internal): update models test * release: 1.75.0 * feat(api): add o3 and o4-mini model IDs * release: 1.74.1 * chore(internal): base client updates * chore(internal): bump pyright version * release: 1.74.0 * feat(api): adding gpt-4.1 family of model IDs * chore(client): minor internal fixes * chore(internal): update pyright settings * fix(chat): skip azure async filter events (#2255) ------------------------------------------------------------------- Thu Apr 17 10:30:36 UTC 2025 - Felix Stegmeier - Update to 1.73.0 * api: add o3 and o4-mini model IDs (4bacbd5) * api: Add evalapi to sdk (#2287) (35262fc) * api: add get /responses/{response_id}/input_items endpoint (4c6a35d) * api: add get /chat/completions endpoint (e6b8a42) * add audio helpers (423655c) * api: new models for TTS, STT, + new audio features for Realtime (#2232) (ab5192d) * api: add support for storing chat completions (#2117) (2357a8f) * client: send X-Stainless-Read-Timeout header (#2094) (0288213) * embeddings: use stdlib array type for improved performance (#2060) (9a95db9) * pagination: avoid fetching when has_more: false (#2098) (1882483) ------------------------------------------------------------------- Fri Jan 3 13:46:26 UTC 2025 - Nico Krapp - Update to 1.59.1: * bump license year (#1981) (f29011a) - Update to 1.59.0: * azure: support for the Realtime API (#1963) (9fda141) * realtime: update docstrings (#1964) (3dee863) ------------------------------------------------------------------- Thu Sep 19 07:33:34 UTC 2024 - Markéta Machová - Add missing dependency ------------------------------------------------------------------- Thu Sep 5 15:50:04 UTC 2024 - Guang Yee - Enable sle15_python_module_pythons. ------------------------------------------------------------------- Wed Aug 21 04:07:02 UTC 2024 - Guang Yee - update to 1.40.8 * types: define FilePurpose enum (#1653) (3c2eeae) ------------------------------------------------------------------- Sat Mar 23 12:06:17 UTC 2024 - Dirk Müller - update to 1.14.2: * cache TypeAdapters * cache TypeAdapters * assistant improvements * fix typo in CONTRIBUTING.md ------------------------------------------------------------------- Sat Mar 16 08:54:28 UTC 2024 - Dirk Müller - update to 1.14.0: * **assistants:** add support for streaming * **streaming:** improve error messages ------------------------------------------------------------------- Mon Mar 11 21:21:09 UTC 2024 - Ben Greiner - Update to 1.13.3 ## Features * api: add wav and pcm to response_format (#1189) (dbd20fc) ## Chores * client: use anyio.sleep instead of asyncio.sleep (#1198) (b6d025b) * internal: bump pyright (#1193) (9202e04) * types: extract run status to a named type (#1178) (249ecbd) - Release 1.13.2 ## Bug Fixes * ci: revert "move github release logic to github app" (#1170) (f1adc2e) - Release 1.13.1 ## Chores * internal: bump rye to v0.24.0 (#1168) (84c4256) - Release 1.13.0 ## Features * api: updates (#1146) (79b7675) ## Bug Fixes * api: remove non-GA instance_id param (#1164) (1abe139) ## Chores * client: use correct accept headers for binary data (#1161) (e536437) - Release 1.12.0 ## Features * api: add timestamp_granularities, add gpt-3.5-turbo-0125 model (#1125) (1ecf8f6) * cli/images: add support for --model arg (#1132) (0d53866) ## Bug Fixes * remove double brackets from timestamp_granularities param (#1140) (3db0222) * types: loosen most List params types to Iterable (#1129) (bdb31a3) ## Chores * internal: add lint command (#1128) (4c021c0) * internal: support serialising iterable types (#1127) (98d4e59) - Restrict to pytest-asyncio < 0.23 - Fix License tag: They changed it for release 1.0.0 gh#openai/openai-python#677 ------------------------------------------------------------------- Fri Jan 26 12:54:53 UTC 2024 - Markéta Machová , Module Updater - Update to 1.10.0 * Auto-retry with backoff if there's an error * Proper types (for mypy/pyright/editors) * You can now instantiate a client, instead of using a global default. * Switch to explicit client instantiation * Weights and Biases CLI will now be included in their own package * many more smaller changes, see https://github.com/openai/openai-python/blob/main/CHANGELOG.md - Use an upstream tarball and run at least some tests ------------------------------------------------------------------- Mon Jul 24 20:18:16 UTC 2023 - Dirk Müller - update to 0.27.8: * Periodically close open `request.Sessions` to avoid buggy interaction with Docker Desktop * Support for Azure Dall-e * Embeddings util - remove unnecessary sort by index * allow api_version to be set by environment variable * Fixed CLI streamed chat completions. * Update README to use gpt-3.5-turbo by default * Updating code and readme to reference the Azure's newest 2023-05-15 API Version * allow for passing requests.Session * fix error messages mentioning support@ * Speed up imports * Add proxy argument support for openai client * Add `__version__` in `__init__.py` * Fix grabbing value out of header * add case to handle indented .json parsing (#175) * Add `openai/datalib/__init__.py` * updated the link for OpenAI API ------------------------------------------------------------------- Thu Jan 26 10:13:45 UTC 2023 - Matej Cepl - Clean up the SPEC file. ------------------------------------------------------------------- Thu Jan 26 07:29:00 UTC 2023 - Adrian Schröter - initial package of version 0.26.2