NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
crates.io
crates.io, #3489 most downloaded on crates.io
Rust library for OpenAI
Latest stable
0.43.0
Last release 21 days ago
01 Nov 2026
Ships fairly regularly
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
1 version withdrawn
withdrawn after publishing
4 years old
127 releases · first in 2022
Drag across the ruler to compare, or tap a release to open it.
feat: spec sync + decisions example by @64bit in #597
Full Changelog: async-openai-v0.42.3...async-openai-v0.43.0
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.42.3...async-openai-v0.43.0
docs: fix configurable request example syntax by @dajiaohuang in #594
Full Changelog: async-openai-v0.42.2...async-openai-v0.42.3
Your coding agent can read these notes before it upgrades. Set up the MCP server →
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.42.2...async-openai-v0.42.3
feat(types): add optional chat and completion schema exports by @GuanLuo in #591
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.42.1...async-openai-v0.42.2
fix(chat): skip_serializing_if on ChatCompletionMessageToolCallChunk optional fields by @Atharva-Kanherkar in #590
Full Changelog: async-openai-v0.42.0...async-openai-v0.42.1
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.42.0...async-openai-v0.42.1
[codex] Add max reasoning effort by @michaelfeil in #573
Full Changelog: async-openai-v0.41.3...async-openai-v0.42.0
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.41.3...async-openai-v0.42.0
fix(types): make FileObject fields public by @coinmoles in #582
FileObject fields public by @coinmoles in #582Full Changelog: async-openai-v0.41.2...async-openai-v0.41.3
FileObject fields public by @coinmoles in https://github.com/64bit/async-openai/pull/582Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.41.2...async-openai-v0.41.3
Exit stream reader task on consumer drop by @max-zheng in #577
Full Changelog: async-openai-v0.41.1...async-openai-v0.41.2
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.41.1...async-openai-v0.41.2
fix(types): omit ImageUrl.detail when None by @yourlogarithm in #564
Full Changelog: async-openai-v0.41.0...async-openai-v0.41.1
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.41.0...async-openai-v0.41.1
feat: upstream spec sync by @64bit in #563
Full Changelog: async-openai-v0.40.3...async-openai-v0.41.0
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.40.3...async-openai-v0.41.0
fix(chat): skip_serializing_if on FunctionCallStream optional fields by @pennycoders in #561
Full Changelog: async-openai-v0.40.2...async-openai-v0.40.3
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.40.2...async-openai-v0.40.3
Surface streaming HTTP errors at the function call instead of as first stream item by @max-zheng in #558
Full Changelog: async-openai-v0.40.1...async-openai-v0.40.2
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.40.1...async-openai-v0.40.2
feat: forward reqwest's rustls-no-provider feature by @valkum in #556
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.40.0...async-openai-v0.40.1
Breaking Change : OpenAIError::ApiError(ApiError) is now OpenAIError:ApiError(ApiErrorResponse)
Breaking Change: OpenAIError::ApiError(ApiError) is now OpenAIError:ApiError(ApiErrorResponse)
Full Changelog: async-openai-v0.39.0...async-openai-v0.40.0
Breaking Change: OpenAIError::ApiError(ApiError) is now OpenAIError:ApiError(ApiErrorResponse)
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.39.0...async-openai-v0.40.0
Only types have breaking changes, if you're not using types from the following PRs, your upgrade should be no-op!
Only types have breaking changes, if you're not using types from the following PRs, your upgrade should be no-op!
action optional in response.output_item.added by @rijkvanzanten in #550Full Changelog: async-openai-v0.38.2...async-openai-v0.39.0
Only types have breaking changes, if you're not using types from the following PRs, your upgrade should be no-op!
action optional in response.output_item.added by @rijkvanzanten in https://github.com/64bit/async-openai/pull/550Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.38.2...async-openai-v0.39.0
fix(responses): default missing InputImageContent.detail during deserialization by @ajcasagrande in #546
Full Changelog: async-openai-v0.38.1...async-openai-v0.38.2
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.38.1...async-openai-v0.38.2
fix: check [DONE] SSE marker before deserialization in stream_mapped_raw_events by @Linuxdazhao in https://github.com/64bit/async-openai/pull/543
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.38.0...async-openai-v0.38.1
fix: always have type in content field of ReasoningItem by @64bit in https://github.com/64bit/async-openai/pull/540
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.37.0...async-openai-v0.38.0
feat: Tower based Middlewares by @64bit in https://github.com/64bit/async-openai/pull/539
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.36.1...async-openai-v0.37.0
From for Item (and InputItem) for multi-turn responses by @Madoshakalaka in https://github.com/64bit/async-openai/pull/538
From<OutputItem> for Item (and InputItem) for multi-turn responses by @Madoshakalaka in https://github.com/64bit/async-openai/pull/538Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.36.0...async-openai-v0.36.1
chore: upgrade reqwest to 0.13 and remove reqwest-eventsource by @richarddd in https://github.com/64bit/async-openai/pull/537
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.35.0...async-openai-v0.36.0
Breaking changes in upstream spec for certain types (diff in PRs), if you're not using any of those types upgrade should be no-op!
Breaking changes in upstream spec for certain types (diff in PRs), if you're not using any of those types upgrade should be no-op!
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.34.0...async-openai-v0.35.0
Breaking changes in upstream spec for certain types (details in PR), if you're not using any of those types upgrade should be no-op!
Breaking changes in upstream spec for certain types (details in PR), if you're not using any of those types upgrade should be no-op!
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.33.1...async-openai-v0.34.0
Implement std::Error for ApiError by @BroderickCarlin in https://github.com/64bit/async-openai/pull/526
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.33.0...async-openai-v0.33.1
A few Rust types have breaking changes from upstream, hence the minor version bump.
A few Rust types have breaking changes from upstream, hence the minor version bump.
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.32.4...async-openai-v0.33.0
feat: sync from upstream spec by @64bit in https://github.com/64bit/async-openai/pull/516
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.32.3...async-openai-v0.32.4
Fix ChatCompletionStreamOptions serialization for OpenAI-compatible providers by @ewgenius in https://github.com/64bit/async-openai/pull/512
ChatCompletionStreamOptions serialization for OpenAI-compatible providers by @ewgenius in https://github.com/64bit/async-openai/pull/512Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.32.2...async-openai-v0.32.3
fix: output_token_details in ImageGenUsage to be Optional by @64bit in https://github.com/64bit/async-openai/pull/508
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.32.1...async-openai-v0.32.2
feat: WASM support by @64bit in https://github.com/64bit/async-openai/pull/506
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.32.0...async-openai-v0.32.1
Updated Types A few types have breaking changes from upstream, hence the minor version bump, if you're not using these types then crate version upgrad…
Added:
Updated Types A few types have breaking changes from upstream, hence the minor version bump, if you're not using these types then crate version upgrade should be no-op:
RealtimeServerEvent has a new variant InputAudioBufferDtmfEventReceivedItem and OutputItem from types/responses/response.rs has updated variantsReasoningEffort has a new variant for xhighAuditLogEventType has updated variantsImageModel has updated variantsCreateVideoRequest is updatedVideoJob -> VideoResource and updated field types, VideoJobError -> VideoResourceError; ListVideoResponse -> VideoListResource with updated fields; VideoJobMetadata -> DeletedVideoResourceexamples/video to use updated typesEvalItemContent and types nested under are updated.Updated Spec Synced from https://app.stainless.com/api/spec/documented/openai/openapi.documented.yml
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.1...async-openai-v0.32.0
Fix type issue with ListFilesResponse by @MakotoE in https://github.com/64bit/async-openai/pull/495
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0...async-openai-v0.31.1
This release sync spec (latest openapi.documented.yml) from upstream to implement all the APIs, along with following features:
This release sync spec (latest openapi.documented.yml) from upstream to implement all the APIs, along with following features:
.path(), .query(), .header(), or .headers() on API group (e.g. client.chat().query()?.list().types::chat, types::responses etc.).response-types, realtime-types etc. for types; types for all types; responses, chat-completion etc. for APIs).byot feature.This is culmination of all the v.0.31.0-alpha. releases!
This is culmination of all the v.0.31.0-alpha.<n> releases!
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.14...async-openai-v0.31.0-beta.1
feat: crate features flags by @64bit in https://github.com/64bit/async-openai/pull/493
responses, chat-completion, etc.) or types ( types, response-types, chat-completion-types, etc.)Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.13...async-openai-v0.31.0-alpha.14
feat: ergonomics, consistency and contained types by @64bit in https://github.com/64bit/async-openai/pull/491
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.12...async-openai-v0.31.0-alpha.13
feat: responses get/retrieve streaming + ergonomics by @64bit in https://github.com/64bit/async-openai/pull/489
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.11...async-openai-v0.31.0-alpha.12
feat: responses & crate ergonomics + bug fixes by @64bit in https://github.com/64bit/async-openai/pull/485
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.10...async-openai-v0.31.0-alpha.11
feat: configurable per request path by @64bit in https://github.com/64bit/async-openai/pull/479
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.9...async-openai-v0.31.0-alpha.10
feat: RequestOptions for individual requests by @64bit in https://github.com/64bit/async-openai/pull/476
client.chat().query(&[("limit", 10)])?.list().await? or client.chat().header("key", "value")?.list().await?)Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.8...async-openai-v0.31.0-alpha.9
feat: spec sync: new tools apply_patch and shell by @64bit in https://github.com/64bit/async-openai/pull/475
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.7...async-openai-v0.31.0-alpha.8
feat: complete spec sync by @64bit in https://github.com/64bit/async-openai/pull/473
feat: complete spec sync by @64bit in https://github.com/64bit/async-openai/pull/473
u64client.admin()Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.6...async-openai-v0.31.0-alpha.7
The alpha releases in v0.31.0 series are intended to sync spec from upstream
The alpha releases in v0.31.0 series are intended to sync spec from upstream
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.5...async-openai-v0.31.0-alpha.6
The alpha releases in v0.31.0 series are intended to sync spec from upstream
The alpha releases in v0.31.0 series are intended to sync spec from upstream
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.4...async-openai-v0.31.0-alpha.5
The alpha releases in v0.31.0 series are intended to sync spec from upstream
The alpha releases in v0.31.0 series are intended to sync spec from upstream
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.3...async-openai-v0.31.0-alpha.4
The alpha releases in v0.31.0 series are intended to sync spec from upstream
The alpha releases in v0.31.0 series are intended to sync spec from upstream
CreateImageEditRequest by @spantaleev in https://github.com/64bit/async-openai/pull/366Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.2...async-openai-v0.31.0-alpha.3
The alpha releases in v0.31.0 series are intended to sync spec from upstream
The alpha releases in v0.31.0 series are intended to sync spec from upstream
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.31.0-alpha.1...async-openai-v0.31.0-alpha.2
The alpha releases in v0.31.0 series are intended to sync spec from upstream, breaking changes are mostly types and their shapes.
The alpha releases in v0.31.0 series are intended to sync spec from upstream, breaking changes are mostly types and their shapes.
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.30.1...async-openai-v0.31.0-alpha.1
feat: video api by @64bit in https://github.com/64bit/async-openai/pull/449
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.30.0...async-openai-v0.30.1
feat: better streaming errors by @64bit in https://github.com/64bit/async-openai/pull/445
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.6...async-openai-v0.30.0
Add support for expires_after field in CreateFileRequest and expires_at field in OpenAIFile by @MakotoE in https://github.com/64bit/async-openai/pull/
name field to ResponseFunctionCallArgumentsDone by @arist76 in https://github.com/64bit/async-openai/pull/442Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.5...async-openai-v0.29.6
feat: Adds file_url to InputFile. Makes text on InputText public. by @lukasfri in https://github.com/64bit/async-openai/pull/441
file_url to InputFile. Makes text on InputText public. by @lukasfri in https://github.com/64bit/async-openai/pull/441Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.4...async-openai-v0.29.5
responses: add verbosity option to TextConfig by @joune in https://github.com/64bit/async-openai/pull/439
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.3...async-openai-v0.29.4
fix: CompoundFilter should use CompoundType instead by @the-spice-must-flow in https://github.com/64bit/async-openai/pull/429
OutputItem to align with OpenAI's Specification by @Advayp in https://github.com/64bit/async-openai/pull/426Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.2...async-openai-v0.29.3
Add skip_serializing_if to more option types by @howardjohn in https://github.com/64bit/async-openai/pull/412
ServiceTier enum for the Responses API by @Advayp in https://github.com/64bit/async-openai/pull/419Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.1...async-openai-v0.29.2
Fix Typo in AssistantStreamEvent Variant (TreadCreated → ThreadCreated) by @posky in https://github.com/64bit/async-openai/pull/400
TreadCreated → ThreadCreated) by @posky in https://github.com/64bit/async-openai/pull/400ChatCompletionToolChoiceOption docs by @wtdcode in https://github.com/64bit/async-openai/pull/401minimal reasoning effort for gpt-5 by @timonv in https://github.com/64bit/async-openai/pull/411Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.29.0...async-openai-v0.29.1
Enable dyn dispatch by dyn Config objects by @ifsheldon in https://github.com/64bit/async-openai/pull/383
CreateResponse and Response structs for Responses API by @paulhendricks in https://github.com/64bit/async-openai/pull/394Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.28.3...async-openai-v0.29.0
Fix: regression with 0.28.2; web search options causes error when using reasoning models; add skip serializing if none to fix. by @spencerbart in http
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.28.2...async-openai-v0.28.3
Implement vector store search, retrieve file content operations by @cfraz89 in https://github.com/64bit/async-openai/pull/360
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.28.1...async-openai-v0.28.2
Add simple ollama chat example by @matthewhaynesonline in https://github.com/64bit/async-openai/pull/336
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.28.0...async-openai-v0.28.1
feat: Bring your own types by @64bit in https://github.com/64bit/async-openai/pull/342
Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.27.2...async-openai-v0.28.0
Add From implementation for ChatCompletionRequestMessage by @keithschulze in https://github.com/64bit/async-openai/pull/318
From<ChatCompletionRequestDeveloperMessage> implementation for ChatCompletionRequestMessage by @keithschulze in https://github.com/64bit/async-openai/pull/318Full Changelog: https://github.com/64bit/async-openai/compare/async-openai-v0.27.1...async-openai-v0.27.2
One column per quarter.