LangChain
LLM application framework. Releases ship breaking changes often, worth watching.
Release history
langchain== 1.4.7
LangChain 1.4.7 adds tool-call-delta stream consumption and LangChainTracer address configuration, plus dependency updates
langchain-core== 1.6.13
LangChain Core now streams tool-call-delta events, enabling real-time partial tool invocation data in stream consumers
langchain-core== 1.6.12
Content-block-start events now propagate extras, fixing event metadata handling in streaming chains
langchain-core== 1.6.11
LangChainTracer now accepts custom addresses for tracing, enabling flexible instrumentation of LLM applications across different deployment environments
langchain== 1.4.6
Adds opt-in per-call interrupts to the human-in-the-loop middleware
langchain-openai== 1.7.2
LangChain OpenAI 1.7.2 fixes streaming response handling when output indices drift across block boundaries
langchain== 1.4.5
Adds summarizer support to SummarizationMiddleware; Check if using middleware-based conversation summarization
langchain-anthropic== 1.7.6
Refreshes model profile data and declares mid-conversation support; Relevant if pinning model profiles explicitly
langchain== 1.4.4
SummarizationMiddleware now retries on context overflow instead of failing, improving reliability for long-form summarization tasks
langchain-openai== 1.7.0
Langchain-openai 1.7.0 adds support for OpenAI's Decisions API, enabling structured decision-making in AI agent workflows
langchain-fireworks== 1.7.1
Fireworks integration now preserves reasoning output during streaming and tool use, fixing a data loss issue in agentic workflows
langchain-huggingface== 1.2.3
Langchain-huggingface 1.2.3 hides API tokens from repr output, switches to torch.accelerator for device detection, and updates dependencies including transformers to 5.10.1
langchain-core== 1.6.7
Bedrock converse now includes OpenAI redacted_content in v1 output; Python 3.14 compatibility hardened
langchain-text-splitters== 1.1.3
Dependency bumps only; No functional change to splitter behavior
langchain-openai== 1.6.7
Discovers Azure OpenAI models; Refreshes model profile data
langchain-anthropic== 1.7.5
Supports Claude Sonnet 5.5 compatibility. Serializes invalid tool calls as tool use on replay
langchain== 1.4.3
Sanitizes cache settings for fallback models, supports Bedrock Mantle chat init, recognizes GPT-6 structured output
langchain-anthropic== 1.7.4
Adds Opus 5.5 and GPT-6 profile augmentations, plus mid-conversation tool changes on SystemMessage
langchain-openai== 1.6.5
Adds Opus 5.5 and GPT-6 profile augmentations, plus mid-conversation tool changes on SystemMessage
langchain-openai== 1.6.4
Refreshes OpenAI model profile data
langchain-deepseek== 1.1.1
Fixes compatible minimum OpenAI dependencies
langchain-openai== 1.6.3
Exposes inferred Responses API routing at initialization, supports GPT-6 request constraints
langchain-typesafe== 0.0
LangChain TypeSafe adds typed routing and classification for AI agents via experimental middleware for safer model selection and tracing
langchain== 1.4.2
Preserves model-generated tool calls in human-in-the-loop tool call edits, adds notice to ToolMessage
langchain-typesafe== 0.0
Initial release with experimental AutoModeMiddleware, ModelRouterMiddleware, and TypeSafeClassifier; Fixes trace usage metadata
langchain-openai== 1.6.2
Adds GPT-6 Astra reasoning efforts and bumps httpx2 to 2.12.0
langchain-openai== 1.6.1
Supports Azure AD auth with OpenAI 3.8; Bumps max completion tokens in cache test
langchain-core== 1.6.2
Adds async tools support and avoids mutation in certain code paths
langchain== 1.4.0
Introduces langchain.mcp namespace and MCPAdapter, omits middleware trace inputs for performance
langchain-anthropic== 1.7.1
Adds Claude Fable 5.1 support, omits middleware trace inputs
langchain== 1.4
LangChain 1.4 adds Model Context Protocol integration with FastMCP 4.0.1, enabling structured tool calling across multiple MCP servers via LangGraph interrupts
langchain== 1.4
LangChain 1.4.0a3 adds langchain.mcp namespace for adapting MCP servers as tools with discovery, caching, and human-in-the-loop interrupt support
langchain-fireworks== 1.6.1
Drops reasoning history blocks
langchain-anthropic== 1.7.0
Adds top-level param for skills via container, updates thinking display mode, supports 1.0 SDK, auto-appends advisor-tool-2026-03-01
langchain-perplexity== 1.4.1
Perplexity integration fixes message type handling, preserves caller parameters, and updates dependencies for compatibility
langchain-openai== 1.6.0
Langchain-openai adds standard model exception types and improves error handling for unexpected OpenAI API response types
langchain-anthropic== 1.6.0
LangChain's Anthropic integration adds standard model exception types and fixes directory search scope issues
langchain-core== 1.6.0
Resolves tool schema serialization issues, adds standard exception types, and improves Windows compatibility in langchain-core
langchain-openai== 1.5.2
Fixes reasoning item boundaries, adds gateway metadata extraction, and extends token counting support to o-series models
langchain-openai== 1.5
Adds gateway metadata extraction from response headers. Fixes o-series token counting
langchain-openrouter== 0.2.8
LangChain OpenRouter now preserves cost metadata and provider information in usage chunks and response metadata
langchain-anthropic== 1.5.6
LangChain Anthropic 1.5.6 fixes tool result block normalization and corrects model profile data for recent Claude versions
langchain== 1.3.15
LangChain 1.3.15 adds state_schema to wrap_tool_call, exposes trace_policy on AgentMiddleware, and fixes middleware control flow, tool call handling, and structured output parsing
langchain-core== 1.5.4
Langchain-core 1.5.4 fixes Pydantic 2.14 compatibility, stops StructuredPrompt from mutating caller kwargs, and preserves tool argument schemas
langchain-openai== 1.4.2
Filters langchain-generated content-block IDs; Handles ContextWindowExceededError
langchain-anthropic== 1.5.4
Langchain-anthropic 1.5.4 fixes tool schema composition handling, preserves tool_choice settings, and adds user_profile_id attribute support
langchain-anthropic== 1.5.2
LangChain Anthropic added Claude Opus 5 model support for AI tooling integrations
langchain-openai== 1.4.1
LangChain OpenAI adds LangSmith gateway support via environment variable and fixes GPT-5.3 profile configuration
langchain-anthropic== 1.5.1
LangChain's Anthropic integration now supports LangSmith gateway via environment variable and enables structured output for Claude Opus 4.8
langchain-anthropic== 1.5.0
Adds reasoning-effort as a standard parameter; Fixes builtin tool recognition
langchain-openai== 1.4.0
Adds reasoning-effort parameter; Bumps Pillow to 12.3.0
langchain-core== 1.5.0
Adds reasoning effort as a standard chat-model parameter
langchain== 1.3.13
LangChain adds Meta model support via new meta extra and implements explicit prompt caching for OpenAI models
langchain-openai== 1.3.4
Suppresses Pydantic serializer warnings on structured output parsed fields and fixes async context loop handling
langchain== 1.3.12
LangChain 1.3.12 fixes interrupt propagation in tool retry middleware, prevents process-group kills in shell operations, and sanitizes Anthropic cache markers on fallback retries
langchain-mistralai== 1.1.6
Surfaces citation metadata from chat responses. Model profile data refreshed
langchain-openrouter== 0.2.6
Adds support for custom HTTP header injection via default headers
langchain== 1.3.11
LangChain 1.3.11 fixes strict tool mode enforcement for non-OpenAI compatible models and updates dependencies
langchain-openai== 1.3.3
Fixes tool schema strictness for OpenAI-compatible models and removes unnecessary response storage fields and stop parameters from API calls
langchain-core== 1.4.8
Langchain-core 1.4.8 memoizes tool schemas, caches JSON schemas, preserves token usage in streaming, and drops Python 3.9 support
langchain-openai== 1.3.1
Fixes streamed tool call normalization and structured output fallbacks, adds package version tracking to tracing metadata
langchain== 1.3.8
Langchain 1.3.8 tightens structured output model fallbacks and fixes async middleware decorator typing
langchain-model-profiles== 0.0.6
Tool call chunk validation added to standard streaming tests, improving LLM integration reliability
langchain-mistralai== 1.1.5
Adds content block token support in callbacks and stabilizes integration tests with rate limiting
langchain-groq== 1.1.3
Langchain-groq 1.1.3 adds content-block-centric streaming v2, strict mode for Groq, standard model properties, and updates dependencies including langsmith 0.8.0 and langchain-core 1.3.3
langchain-openai== 1.3.0
Langchain-openai 1.3.0 adds support for OpenAI's apply_patch built-in tool and updates model profile data
langchain-core== 1.4.2
LangChain Core deprecated the problematic dict() method to encourage use of safer serialization patterns in LLM application chains
langchain-core== 1.4.1
Removes Bedrock prevalidation from load path; Bumps uuid-utils to 0.16.0
langchain-deepseek== 1.1.0
Content-block streaming v2 support, dependency updates including langchain-core 1.3.3, and Azure endpoint handling fixes
langchain== 1.3.4
LangChain 1.3.4 improves human-in-the-loop rejection guidance for better control flow in AI agent workflows
langchain== 1.3.3
LangChain 1.3.3 adds subagent run projection to typed channels and human-in-the-loop interrupt predicates for agentic workflows
langchain-anthropic== 1.4.4
Normalizes cross-provider tool-call IDs in Anthropic integration to ensure consistent behavior across different LLM providers
langchain-fireworks== 1.4.2
Fireworks integration strips non-wire keys from content parts before serialization to prevent serialization errors
langchain-perplexity== 1.3.0
ChatPerplexity now respects responses API flag; No impact unless that parameter is in use
langchain-tests== 1.1.9
Standard test suite relaxes streaming assertion validation to allow extra content blocks in LLM responses
langchain-fireworks== 1.4.0
Migrates to fireworks-ai 1.x SDK. Breaking change for anyone pinning older SDK versions
langchain-tests== 1.1.8
Langchain-tests 1.1.8 constrains langchain-core version bounds and fixes per-call model override behavior in standard tests
langchain== 1.3.0
LangChain agents now support v3 event streaming for improved visibility into agentic execution flows
langchain== 1.2.18
LangChain 1.2.18 reverts agent-type tagging, deprecates hub operations, and retargets create_agent deprecations