langchain-openai
A core integration package in the LangChain ecosystem for connecting to OpenAI models. It provides developers with a standardized interface for building LLM applications on top of the OpenAI API, supporting capabilities such as streaming and tool calling.
Core Facts
Timeline (last 90 days)
langchain-openai 1.7.2 fixes the issue in the OpenAI Responses API where, when output_index drifts in streaming scenarios, streamed return items are incorrectly split across multiple blocks (PR #41121).
After the fix, langchain-openai will recognize output_index drift and proactively merge, keeping the number of blocks seen downstream consistent with the semantic boundaries.
The changes in langchain-openai 1.7.2 since 1.7.1 include release(openai): 1.7.2 (PR #41245) and fix(openai): keep Responses stream items on one block when output_index drifts (PR #41121).
langchain-openai is the adaptation layer between LangChain and OpenAI models, and its release cadence typically follows changes in the upstream API.
For projects already locked to 1.7.1, upgrading to 1.7.2 carries low risk, with a small change scope primarily centered on the fix of one known defect.
For projects using a langchain-openai version below 1.7.1 and involving streaming output scenarios, it is recommended to evaluate upgrading to 1.7.2, with the upgrade command being pip install -U langchain-openai==1.7.2.
The problems caused by output_index drift may manifest as truncated output content, garbled message block concatenation, or in some cases trigger parsing exceptions.
This declaration of mid-conversation support involves three packages at the same time: core, anthropic and openai
The title of PR #41180 is "declare mid-conversation support in model profiles", i.e., declaring mid-conversation support capability in model profiles.
These 1.7.1 release notes do not include detailed behavioral documentation of the mid-conversation support capability, so it is recommended to verify against the target model before enabling it in an actual project.
24 more timeline events
All Facts (20)
The upgrade can be completed via the command pip install -U langchain-core==1.6.6.
90%VerifiedEven for patch updates, it is recommended to run the full test suite in the CI pipeline to ensure consistent application behavior after upgrading.
75%UnverifiedAfter the fix, langchain-openai will recognize output_index drift and proactively merge, keeping the number of blocks seen downstream consistent with the semantic boundaries.
50%Unverifiedlangchain-openai 1.7.2 fixes the issue in the OpenAI Responses API where, when output_index drifts in streaming scenarios, streamed return items are incorrectly split across multiple blocks (PR #41121).
50%UnverifiedThe problems caused by output_index drift may manifest as truncated output content, garbled message block concatenation, or in some cases trigger parsing exceptions.
50%UnverifiedFor projects using a langchain-openai version below 1.7.1 and involving streaming output scenarios, it is recommended to evaluate upgrading to 1.7.2, with the upgrade command being pip install -U langchain-openai==1.7.2.
50%UnverifiedFor projects already locked to 1.7.1, upgrading to 1.7.2 carries low risk, with a small change scope primarily centered on the fix of one known defect.
50%Unverifiedlangchain-openai is the adaptation layer between LangChain and OpenAI models, and its release cadence typically follows changes in the upstream API.
50%UnverifiedThe changes in langchain-openai 1.7.2 since 1.7.1 include release(openai): 1.7.2 (PR #41245) and fix(openai): keep Responses stream items on one block when output_index drifts (PR #41121).
50%UnverifiedIf a project has langchain-core, langchain-openai, langchain-anthropic and other packages installed at the same time, it is recommended to upgrade them all to mutually compatible versions to avoid inconsistency between capability declarations and actual implementations.
50%UnverifiedIf only langchain-openai is upgraded without upgrading core in sync, the upper-layer logic reading profiles may not recognize the new fields, or the declaration may be ignored without error.
50%Unverifiedlangchain-openai 1.7.0 added support for the Decisions API via PR #41125.
50%UnverifiedThe dependency upgrades are concentrated primarily in the /libs/partners/openai directory.
50%UnverifiedThe mid-conversation tool-change feature in version 1.6.5 can simplify the workaround code that developers previously wrote to implement dynamic tool sets
50%UnverifiedIn earlier versions, error events returned by the server during streaming calls to the OpenAI API might not be correctly raised as exceptions, causing errors to be silently swallowed
50%Unverified1.6.6 includes the routine release commit PR #40800 'release(openai): 1.6.6'
50%Unverified1.6.6 is a patch-level version involving only error-handling logic fixes rather than API changes, and is backward compatible
50%UnverifiedThe author recommends that teams using langchain-openai's streaming capabilities evaluate upgrading to 1.6.6 for more predictable exception behavior
50%Unverifiedlangchain-openai is the core integration component in the LangChain ecosystem for connecting to OpenAI models
50%UnverifiedThe core fix in 1.6.6 is PR #40791 'fix(openai): raise on error events in stream path'
50%