langchain-core
Description
🦜🍎️ LangChain Core
Looking for the JS/TS version? Check out LangChain.js.
To help you ship LangChain apps to production faster, check out LangSmith. LangSmith is a unified developer platform for building, testing, and monitoring LLM applications.
Quick Install
uv add langchain-core🤔 What is this?
LangChain Core contains the base abstractions that power the LangChain ecosystem.
These abstractions are designed to be as modular and simple as possible.
The benefit of having these abstractions is that any provider can implement the required interface and then easily be used in the rest of the LangChain ecosystem.
⛰️ Why build on top of LangChain Core?
The LangChain ecosystem is built on top of langchain-core. Some of the benefits:
- Modularity: We've designed Core around abstractions that are independent of each other, and not tied to any specific model provider.
- Stability: We are committed to a stable versioning scheme, and will communicate any breaking changes with advance notice and version bumps.
- Battle-tested: Core components have the largest install base in the LLM ecosystem, and are used in production by many companies.
📖 Documentation
For full documentation, see the API reference. For conceptual guides, tutorials, and examples on using LangChain, see the LangChain Docs. You can also chat with the docs using Chat LangChain.
📕 Releases & Versioning
See our Releases and Versioning policies.
💁 Contributing
As an open-source project in a rapidly developing field, we are extremely open to contributions, whether it be in the form of a new feature, improved infrastructure, or better documentation.
For detailed information on how to contribute, see the Contributing Guide.
Resources
- LangChain Academy — comprehensive, free courses on LangChain libraries and products, made by the LangChain team
- Code of Conduct — community guidelines and standards
Classes
Class
BaseRateLimiter
Base class for rate limiters.
Class
InMemoryRateLimiter
An in memory rate limiter based on a token bucket algorithm.
Class
LangSmithRetrieverParams
LangSmith parameters for tracing.
Class
BaseRetriever
Abstract base class for a document retrieval system.
Class
BaseStore
Abstract interface for a key-value store.
Class
InMemoryBaseStore
In-memory implementation of the BaseStore using a dictionary.
Class
InMemoryStore
In-memory store for any type of data.
Class
InMemoryByteStore
In-memory store for bytes.
Class
InvalidKeyException
Raised when a key is invalid; e.g., uses incorrect characters.
Class
LangChainException
General LangChain exception.
Class
TracerException
Base class for exceptions in tracers module.
Class
OutputParserException
Exception that output parsers should raise to signify a parsing error.
Class
ContextOverflowError
Exception raised when input exceeds the model's context limit.
Class
ErrorCode
Error codes.
Class
PromptValue
Base abstract class for inputs to any language model.
Class
StringPromptValue
String prompt value.
Class
ChatPromptValue
Chat prompt value.
Class
ImageURL
Image URL for multimodal model inputs (OpenAI format).
Class
ImagePromptValue
Image prompt value.
Class
ChatPromptValueConcrete
Chat prompt value which explicitly lists out the message types it accepts.
Class
AgentAction
Represents a request to execute an action by an agent.
Class
AgentActionMessageLog
Representation of an action to be executed by an agent.
Class
AgentStep
Result of running an AgentAction.
Class
AgentFinish
Final return value of an ActionAgent.
Class
BaseCrossEncoder
Interface for cross encoder models.
Class
BaseChatLoader
Base class for chat loaders.
Class
Visitor
Defines interface for IR translation using a visitor pattern.
Class
Expr
Base class for all expressions.
Class
Operator
Enumerator of the operations.
Class
Comparator
Enumerator of the comparison operators.
Class
FilterDirective
Filtering expression.
Class
Comparison
Comparison to a value.
Class
Operation
Logical operation over other directives.
Class
StructuredQuery
Structured query.
Class
ChatSession
Chat Session.
Class
BaseChatMessageHistory
Abstract base class for storing chat message history.
Class
InMemoryChatMessageHistory
In memory implementation of chat message history.
Class
BaseCache
Interface for a caching layer for LLMs and Chat models.
Class
InMemoryCache
Cache that stores things in memory.
Class
BaseDocumentTransformer
Abstract base class for document transformation.
Class
BaseDocumentCompressor
Base class for document compressors.
Class
BaseMedia
Base class for content used in retrieval and data processing workflows.
Class
Blob
Raw data abstraction for document loading and file processing.
Class
Document
Class for storing a piece of text and associated metadata.
Class
FakeListLLM
Fake LLM for testing purposes.
Class
FakeListLLMError
Fake error for testing purposes.
Class
FakeStreamingListLLM
Fake streaming list LLM for testing purposes.
Class
SyncProjection
Sync iterable of deltas with pull-based backpressure.
Class
SyncTextProjection
String-specialized sync projection.
Class
AsyncProjection
Async iterable of deltas that is also awaitable for the final value.
Class
ChatModelStream
Synchronous per-message streaming object for a single LLM response.
Class
AsyncChatModelStream
Asynchronous per-message streaming object for a single LLM response.
Class
BaseChatModel
Base class for chat models.
Class
SimpleChatModel
Simplified implementation for a chat model to inherit from.
Class
FakeMessagesListChatModel
Fake chat model for testing purposes.
Class
FakeListChatModelError
Fake error for testing purposes.
Class
FakeListChatModel
Fake chat model for testing purposes.
Class
FakeChatModel
Fake Chat Model wrapper for testing purposes.
Class
GenericFakeChatModel
Generic fake chat model that can be used to test the chat model interface.
Class
ParrotFakeChatModel
Generic fake chat model that can be used to test the chat model interface.
Class
ModelProfile
Description of a chat model's capabilities, exposed via model.profile.
Class
ParsedDataUri
Class
BaseLLM
Base LLM abstract interface.
Class
LLM
Simple interface for implementing a custom LLM.
Class
LangSmithParams
LangSmith parameters for tracing.
Class
BaseLanguageModel
Abstract base class for interfacing with language models.
Class
InMemoryVectorStore
In-memory vector store implementation.
Class
VectorStore
Interface for vector store.
Class
VectorStoreRetriever
Base Retriever class for VectorStore.
Class
LangSmithLoader
Load LangSmith Dataset examples as Document objects.
Class
BlobLoader
Abstract interface for blob loaders implementation.
Class
BaseLoader
Interface for document loader.
Class
BaseBlobParser
Abstract interface for blob parsers.
Class
ChevronError
Custom exception for Chevron errors.
Class
NoLock
Dummy lock that provides the proper interface but no protection.
Class
Tee
Create n separate asynchronous iterators over iterable.
Class
FunctionDescription
Representation of a callable function to send to an LLM.
Class
ToolDescription
Representation of a callable function to the OpenAI API.
Class
NoLock
Dummy lock that provides the proper interface but no protection.
Class
Tee
Create n separate asynchronous iterators over iterable.
Class
aclosing
Async context manager to wrap an AsyncGenerator that has a aclose() method.
Class
GatewayConfig
Resolved gateway configuration.
Class
StrictFormatter
A string formatter that enforces keyword-only argument substitution.
Class
LangChainBetaWarning
A class for issuing beta warnings for LangChain users.
Class
LangChainDeprecationWarning
A class for issuing deprecation warnings for LangChain users.
Class
LangChainPendingDeprecationWarning
A class for issuing deprecation warnings for LangChain users.
Class
UsageMetadataCallbackHandler
Callback Handler that tracks AIMessage.usage_metadata.
Class
BaseRunManager
Base class for run manager (a bound callback manager).
Class
RunManager
Synchronous run manager.
Class
ParentRunManager
Synchronous parent run manager.
Class
AsyncRunManager
Async run manager.
Class
AsyncParentRunManager
Async parent run manager.
Class
CallbackManagerForLLMRun
Callback manager for LLM run.
Class
AsyncCallbackManagerForLLMRun
Async callback manager for LLM run.
Class
CallbackManagerForChainRun
Callback manager for chain run.
Class
AsyncCallbackManagerForChainRun
Async callback manager for chain run.
Class
CallbackManagerForToolRun
Callback manager for tool run.
Class
AsyncCallbackManagerForToolRun
Async callback manager for tool run.
Class
CallbackManagerForRetrieverRun
Callback manager for retriever run.
Class
AsyncCallbackManagerForRetrieverRun
Async callback manager for retriever run.
Class
CallbackManager
Callback manager for LangChain.
Class
CallbackManagerForChainGroup
Callback manager for the chain group.
Class
AsyncCallbackManager
Async callback manager that handles callbacks from LangChain.
Class
AsyncCallbackManagerForChainGroup
Async callback manager for the chain group.
Class
StreamingStdOutCallbackHandler
Callback handler for streaming.
Class
StdOutCallbackHandler
Callback handler that prints to std out.
Class
FileCallbackHandler
Callback handler that writes to a file.
Class
RetrieverManagerMixin
Mixin for Retriever callbacks.
Class
LLMManagerMixin
Mixin for LLM callbacks.
Class
ChainManagerMixin
Mixin for chain callbacks.
Class
ToolManagerMixin
Mixin for tool callbacks.
Class
CallbackManagerMixin
Mixin for callback manager.
Class
RunManagerMixin
Mixin for run manager.
Class
BaseCallbackHandler
Base callback handler.
Class
AsyncCallbackHandler
Base async callback handler.
Class
BaseCallbackManager
Base callback manager.
Class
ChatGeneration
A single chat generation output.
Class
ChatGenerationChunk
ChatGeneration chunk.
Class
ChatResult
Use to represent the result of a chat model call with a single prompt.
Class
Generation
A single text generation output.
Class
GenerationChunk
GenerationChunk, which can be concatenated with other Generation chunks.
Class
LLMResult
A container for results of an LLM call.
Class
RunInfo
Class that contains metadata for a single execution of a chain or model.
Class
RemoveMessage
Message responsible for deleting other messages.
Class
ChatMessage
Message that can be assigned an arbitrary speaker (i.e. role).
Class
ChatMessageChunk
Chat Message chunk.
Class
SystemMessage
Message for priming AI behavior.
Class
SystemMessageChunk
System Message chunk.
Class
InputTokenDetails
Breakdown of input token counts.
Class
OutputTokenDetails
Breakdown of output token counts.
Class
UsageMetadata
Usage metadata for a message, such as token counts.
Class
AIMessage
Message from an AI.
Class
AIMessageChunk
Message chunk from an AI (yielded when streaming).
Class
HumanMessage
Message from the user.
Class
HumanMessageChunk
Human Message chunk.
Class
ToolOutputMixin
Mixin for objects that tools can return directly.
Class
ToolMessage
Message for passing the result of executing a tool back to a model.
Class
ToolMessageChunk
Tool Message chunk.
Class
ToolCall
Represents an AI's request to call a tool.
Class
ToolCallChunk
A chunk of a tool call (yielded when streaming).
Class
Citation
Annotation for citing data from a document.
Class
NonStandardAnnotation
Provider-specific annotation format.
Class
TextContentBlock
Text output from a LLM.
Class
ToolCall
Represents an AI's request to call a tool.
Class
ToolCallChunk
A chunk of a tool call (yielded when streaming).
Class
InvalidToolCall
Allowance for errors made by LLM.
Class
ServerToolCall
Tool call that is executed server-side.
Class
ServerToolCallChunk
A chunk of a server-side tool call (yielded when streaming).
Class
ServerToolResult
Result of a server-side tool call.
Class
ReasoningContentBlock
Reasoning output from a LLM.
Class
ImageContentBlock
Image data.
Class
VideoContentBlock
Video data.
Class
AudioContentBlock
Audio data.
Class
PlainTextContentBlock
Plaintext data (e.g., from a .txt or .md document).
Class
FileContentBlock
File data that doesn't fit into other multimodal block types.
Class
NonStandardContentBlock
Provider-specific content data.
Class
FunctionMessage
Message for passing the result of executing a tool back to a model.
Class
FunctionMessageChunk
Function Message chunk.
Class
TextAccessor
String-like object that supports both property and method access patterns.
Class
BaseMessage
Base abstract message class.
Class
BaseMessageChunk
Message chunk, which can be concatenated with other Message chunks.
Class
SSRFPolicy
Immutable policy controlling which URLs/IPs are considered safe.
Class
SSRFSafeTransport
httpx async transport that validates DNS results against an SSRF policy.
Class
SSRFSafeSyncTransport
httpx sync transport that validates DNS results against an SSRF policy.
Class
SSRFBlockedError
Raised when a request is blocked by SSRF protection policy.
Class
DictPromptTemplate
Template represented by a dictionary.
Class
MessagesPlaceholder
Prompt template that assumes variable is already list of messages.
Class
BaseStringMessagePromptTemplate
Base class for message prompt templates that use a string prompt template.
Class
ChatMessagePromptTemplate
Chat message prompt template.
Class
HumanMessagePromptTemplate
Human message prompt template.
Class
AIMessagePromptTemplate
AI message prompt template.
Class
SystemMessagePromptTemplate
System message prompt template.
Class
BaseChatPromptTemplate
Base class for chat prompt templates.
Class
ChatPromptTemplate
Prompt template for chat models.
Class
StringPromptTemplate
String prompt that exposes the format method, returning a prompt.
Class
FewShotPromptWithTemplates
Prompt template that contains few shot examples.
Class
FewShotPromptTemplate
Prompt template that contains few shot examples.
Class
FewShotChatMessagePromptTemplate
Chat prompt template that supports few-shot examples.
Class
BaseMessagePromptTemplate
Base class for message prompt templates.
Class
PromptTemplate
Prompt template for a language model.
Class
StructuredPrompt
Structured prompt template for a language model.
Class
BasePromptTemplate
Base class for all prompt templates, returning a prompt.
Class
ImagePromptTemplate
Image prompt template for a multimodal model.
Class
LengthBasedExampleSelector
Select examples based on length.
Class
SemanticSimilarityExampleSelector
Select examples based on semantic similarity.
Class
MaxMarginalRelevanceExampleSelector
Select examples based on Max Marginal Relevance.
Class
BaseExampleSelector
Interface for selecting examples to include in prompts.
Class
RunInfo
Information about a run.
Class
EvaluatorCallbackHandler
Tracer that runs a run evaluator whenever a run is persisted.
Class
RunCollectorCallbackHandler
Tracer that collects all nested runs in a list.
Class
LogEntry
A single entry in the run log.
Class
RunState
State of the run.
Class
RunLogPatch
Patch to the run log.
Class
RunLog
Run log.
Class
LogStreamCallbackHandler
Tracer that streams run logs to a stream.
Class
LangChainTracer
Implementation of the SharedTracer that POSTS to the LangChain endpoint.
Class
FunctionCallbackHandler
Tracer that calls a function with a single str parameter.
Class
ConsoleCallbackHandler
Tracer that prints to the console.
Class
RootListenersTracer
Tracer that calls listeners on run start, end, and error.
Class
AsyncRootListenersTracer
Async tracer that calls listeners on run start, end, and error.
Class
BaseTracer
Base interface for tracers.
Class
AsyncBaseTracer
Async base interface for tracers.
Class
Embeddings
Interface for embedding models.
Class
FakeEmbeddings
Fake embedding model for unit testing purposes.
Class
DeterministicFakeEmbedding
Deterministic fake embedding model for unit testing purposes.
Class
IndexingException
Raised when an indexing operation fails.
Class
IndexingResult
Return a detailed a breakdown of the result of the indexing operation.
Class
InMemoryDocumentIndex
In memory document index.
Class
RecordManager
Abstract base class representing the interface for a record manager.
Class
InMemoryRecordManager
An in-memory record manager for testing purposes.
Class
UpsertResponse
A generic response for upsert operations.
Class
DeleteResponse
A generic response for delete operation.
Class
DocumentIndex
A document retriever that supports indexing operations.
Class
Tool
Tool that takes in function or coroutine directly.
Class
StructuredTool
Tool that can operate on any number of inputs.
Class
RetrieverInput
Input to the retriever.
Class
SchemaAnnotationError
Raised when args_schema is missing or has an incorrect type annotation.
Class
ToolException
Exception thrown when a tool execution error occurs.
Class
BaseTool
Base class for all LangChain tools.
Class
InjectedToolArg
Annotation for tool arguments that are injected at runtime.
Class
InjectedToolCallId
Annotation for injecting the tool call ID.
Class
BaseToolkit
Base class for toolkits containing related tools.
Class
Reviver
Reviver for JSON objects.
Class
BaseSerialized
Base class for serialized objects.
Class
SerializedConstructor
Serialized constructor.
Class
SerializedSecret
Serialized secret.
Class
SerializedNotImplemented
Serialized not implemented.
Class
Serializable
Serializable base class.
Class
EventData
Data associated with a streaming event.
Class
BaseStreamEvent
Streaming event.
Class
StandardStreamEvent
A standard stream event that follows LangChain convention for event data.
Class
CustomStreamEvent
Custom stream event created by the user.
Class
DynamicRunnable
Serializable Runnable that can be dynamically configured.
Class
RunnableConfigurableFields
Runnable that can be dynamically configured.
Class
StrEnum
String enum.
Class
RunnableConfigurableAlternatives
Runnable that can be dynamically configured.
Class
VertexViewer
VertexViewer class.
Class
AsciiCanvas
Class for drawing in ASCII.
Class
PngDrawer
Helper class to draw a state graph into a PNG file.
Class
RunnableWithFallbacks
Runnable that can fallback to other Runnable objects if it fails.
Class
EmptyDict
Empty dict type.
Class
RunnableConfig
Configuration for a Runnable.
Class
ContextThreadPoolExecutor
ThreadPoolExecutor that copies the context to the child thread.
Class
RunnableWithMessageHistory
Runnable that manages chat message history for another Runnable.
Class
RunnableBranch
Runnable that selects which branch to run based on a condition.
Class
ExponentialJitterParams
Parameters for tenacity.wait_exponential_jitter.
Class
RunnableRetry
Retry a Runnable if it fails.
Class
RunnablePassthrough
Runnable to passthrough inputs unchanged or with additional keys.
Class
RunnableAssign
Runnable that assigns key-value pairs to dict[str, Any] inputs.
Class
RunnablePick
Runnable that picks keys from dict[str, Any] inputs.
Class
Stringifiable
Protocol for objects that can be converted to a string.
Class
LabelsDict
Dictionary of labels for nodes and edges in a graph.
Class
Edge
Edge in a graph.
Class
Node
Node in a graph.
Class
Branch
Branch in a graph.
Class
CurveStyle
Enum for different curve styles supported by Mermaid.
Class
NodeStyles
Schema for Hexadecimal color codes for different node types.
Class
MermaidDrawMethod
Enum for different draw methods supported by Mermaid.
Class
Graph
Graph of nodes and edges.
Class
RouterInput
Router input.
Class
RouterRunnable
Runnable that routes to a set of Runnable based on Input['key'].
Class
Runnable
A unit of work that can be invoked, batched, streamed, transformed and composed.
Class
RunnableSerializable
Runnable that can be serialized to JSON.
Class
RunnableSequence
Sequence of Runnable objects, where the output of one is the input of the next.
Class
RunnableParallel
Runnable that runs a mapping of Runnables in parallel.
Class
RunnableGenerator
Runnable that runs a generator function.
Class
RunnableLambda
RunnableLambda converts a python callable into a Runnable.
Class
RunnableEachBase
RunnableEachBase class.
Class
RunnableEach
RunnableEach class.
Class
RunnableBindingBase
Runnable that delegates calls to another Runnable with a set of **kwargs.
Class
RunnableBinding
Wrap a Runnable with additional functionality.
Class
IsLocalDict
Check if a name is a local dict.
Class
IsFunctionArgDict
Check if the first argument of a function is a dict.
Class
NonLocals
Get nonlocal variables accessed.
Class
FunctionNonLocals
Get the nonlocal variables accessed of a function.
Class
GetLambdaSource
Get the source code of a lambda function.
Class
AddableDict
Dictionary that can be added to another dictionary.
Class
SupportsAdd
Protocol for objects that support addition.
Class
ConfigurableField
Field that can be configured by the user.
Class
ConfigurableFieldSingleOption
Field that can be configured by the user with a default value.
Class
ConfigurableFieldMultiOption
Field that can be configured by the user with multiple default values.
Class
ConfigurableFieldSpec
Field that can be configured by the user. It is a specification of a field.
Class
XMLOutputParser
Parse an output using xml format.
Class
OutputFunctionsParser
Parse an output that is one of sets of values.
Class
JsonOutputFunctionsParser
Parse an output as the JSON object.
Class
JsonKeyOutputFunctionsParser
Parse an output as the element of the JSON object.
Class
PydanticOutputFunctionsParser
Parse an output as a Pydantic object.
Class
PydanticAttrOutputFunctionsParser
Parse an output as an attribute of a Pydantic object.
Class
StrOutputParser
Extract text content from model outputs as a string.
Class
PydanticOutputParser
Parse an output using a Pydantic model.
Class
ListOutputParser
Parse the output of a model to a list.
Class
CommaSeparatedListOutputParser
Parse the output of a model to a comma-separated list.
Class
NumberedListOutputParser
Parse a numbered list.
Class
MarkdownListOutputParser
Parse a Markdown list.
Class
JsonOutputToolsParser
Parse tools from OpenAI response.
Class
JsonOutputKeyToolsParser
Parse tools from OpenAI response.
Class
PydanticToolsParser
Parse tools from OpenAI response.
Class
JsonOutputParser
Parse the output of an LLM call to a JSON object.
Class
BaseLLMOutputParser
Abstract base class for parsing the outputs of a model.
Class
BaseGenerationOutputParser
Base class to parse the output of an LLM call.
Class
BaseOutputParser
Base class to parse the output of an LLM call.
Class
BaseTransformOutputParser
Base class for an output parser that can handle streaming input.
Class
BaseCumulativeTransformOutputParser
Base class for an output parser that can handle streaming input.
Functions
Function
set_verbose
Set a new value for the verbose global setting.
Function
get_verbose
Get the value of the verbose global setting.
Function
set_debug
Set a new value for the debug global setting.
Function
get_debug
Get the value of the debug global setting.
Function
set_llm_cache
Set a new LLM cache, overwriting the previous value, if any.
Function
get_llm_cache
Get the value of the llm_cache global setting.
Function
print_sys_info
Print information about the environment for debugging purposes.
Function
import_attr
Import an attribute from a module located in a package.
Function
create_message
Create a message with a link to the LangChain troubleshooting guide.
Function
get_runtime_environment
Get information about the LangChain runtime environment.
Function
finalize_tool_call_chunk
Parse accumulated tool-chunk args into a finalized block.
Function
chunks_to_events
Convert a stream of ChatGenerationChunk to protocol events.
Function
achunks_to_events
Async variant of chunks_to_events.
Function
message_to_events
Replay a finalized message as a synthetic event lifecycle.
Function
amessage_to_events
Async variant of message_to_events.
Function
generate_from_stream
Generate from a stream.
Function
agenerate_from_stream
Async generate from a stream.
Function
is_openai_data_block
Check whether a block contains multimodal data in OpenAI Chat Completions format.
Function
create_base_retry_decorator
Create a retry decorator for a given LLM and provided a list of error types.
Function
get_prompts
Get prompts that are already cached.
Function
aget_prompts
Get prompts that are already cached. Async version.
Function
update_cache
Update the cache and get the LLM output.
Function
aupdate_cache
Update the cache and get the LLM output. Async version.
Function
get_tokenizer
Get a GPT-2 tokenizer instance.
Function
maximal_marginal_relevance
Calculate maximal marginal relevance.
Function
grab_literal
Parse a literal from the template.
Function
l_sa_check
Do a preliminary check to see if a tag could be a standalone.
Function
r_sa_check
Do a final check to see if a tag could be a standalone.
Function
parse_tag
Parse a tag from a template.
Function
tokenize
Tokenize a mustache template.
Function
render
Render a mustache template.
Function
get_color_mapping
Get mapping for items to a support color.
Function
get_colored_text
Get colored text.
Function
get_bolded_text
Get bolded text.
Function
print_text
Print text with highlighting and no end characters.
Function
tee_peer
An individual iterator of a .tee.
Function
batch_iterate
Utility batching function.
Function
dereference_refs
Resolve and inline JSON Schema $ref references in a schema object.
Function
find_all_links
Extract all links from a raw HTML string.
Function
extract_sub_links
Extract all links from a raw HTML string and convert into absolute paths.
Function
uuid7
Generate a UUID from a Unix timestamp in nanoseconds and random bits.
Function
convert_to_openai_function
Convert a raw function/class to an OpenAI function.
Function
convert_to_openai_tool
Convert a tool-like object to an OpenAI tool schema.
Function
convert_to_json_schema
Convert a schema representation to a JSON schema.
Function
tool_example_to_messages
Convert an example into a list of messages that can be fed into an LLM.
Function
tee_peer
An individual iterator of a tee.
Function
abatch_iterate
Utility batching function for async iterables.
Function
is_pydantic_v1_subclass
Check if the given class is Pydantic v1-like.
Function
is_pydantic_v2_subclass
Check if the given class is Pydantic v2-like.
Function
is_basemodel_subclass
Check if the given class is a subclass of Pydantic BaseModel.
Function
is_basemodel_instance
Check if the given class is an instance of Pydantic BaseModel.
Function
pre_init
Decorator to run a function before model initialization.
Function
get_fields
Return the field names of a Pydantic model.
Function
model_json_schema
Return the JSON schema of a Pydantic model class of either major version.
Function
model_validate
Validate obj against a Pydantic model class of either major version.
Function
create_model
Create a Pydantic model with the given field definitions.
Function
create_model_v2
Create a Pydantic model with the given field definitions.
Function
parse_partial_json
Parse a JSON string that may be missing closing braces.
Function
parse_json_markdown
Parse a JSON string from a Markdown string.
Function
parse_and_check_json_markdown
Parse and check a JSON string from a Markdown string.
Function
stringify_value
Stringify a value.
Function
stringify_dict
Stringify a dictionary.
Function
comma_list
Convert an iterable to a comma-separated string.
Function
sanitize_for_postgres
Sanitize text by removing NUL bytes that are incompatible with PostgreSQL.
Function
env_var_is_set
Check if an environment variable is set.
Function
get_from_dict_or_env
Get a value from a dictionary or an environment variable.
Function
get_from_env
Get a value from a dictionary or an environment variable.
Function
is_interactive_env
Determine if running within IPython or Jupyter.
Function
xor_args
Validate specified keyword args are mutually exclusive.
Function
raise_for_status_with_text
Raise an error with the response text.
Function
mock_now
Context manager for mocking out datetime.now() in unit tests.
Function
guard_import
Dynamically import a module.
Function
check_package_version
Check the version of a package.
Function
get_pydantic_field_names
Get field names, including aliases, for a pydantic class.
Function
build_extra_kwargs
Build extra kwargs from values and extra_kwargs.
Function
convert_to_secret_str
Convert a string to a SecretStr if needed.
Function
from_env
Create a factory method that gets a value from an environment variable.
Function
secret_from_env
Secret from env.
Function
ensure_id
Ensure the ID is a valid string, generating a new UUID if not provided.
Function
merge_dicts
Merge dictionaries.
Function
merge_lists
Add many lists, handling None.
Function
merge_obj
Merge two objects.
Function
beta
Decorator to mark a function, a class, or a property as beta.
Function
suppress_langchain_beta_warning
Context manager to suppress LangChainDeprecationWarning.
Function
warn_beta
Display a standardized beta annotation.
Function
surface_langchain_beta_warnings
Unmute LangChain beta warnings.
Function
deprecated
Decorator to mark a function, a class, or a property as deprecated.
Function
suppress_langchain_deprecation_warning
Context manager to suppress LangChainDeprecationWarning.
Function
warn_deprecated
Display a standardized deprecation.
Function
surface_langchain_deprecation_warnings
Unmute LangChain deprecation warnings.
Function
rename_parameter
Decorator indicating that parameter old of func is renamed to new.
Function
get_relative_path
Get the path of the file as a relative path to the package directory.
Function
as_import_path
Path of the file as a LangChain import exclude langchain top namespace.
Function
is_caller_internal
Return whether the caller at depth of this function is internal.
Function
get_usage_metadata_callback
Get usage metadata callback.
Function
trace_as_chain_group
Get a callback manager for a chain group in a context manager.
Function
atrace_as_chain_group
Get an async callback manager for a chain group in a context manager.
Function
shielded
Makes so an awaitable method is always shielded from cancellation.
Function
handle_event
Generic event handler for CallbackManager.
Function
ahandle_event
Async generic event handler for AsyncCallbackManager.
Function
adispatch_custom_event
Dispatch an adhoc event to the handlers.
Function
dispatch_custom_event
Dispatch an adhoc event.
Function
merge_chat_generation_chunks
Merge a list of ChatGenerationChunks into a single ChatGenerationChunk.
Function
add_ai_message_chunks
Add multiple AIMessageChunks together.
Function
add_usage
Recursively add two UsageMetadata objects.
Function
subtract_usage
Recursively subtract two UsageMetadata objects.
Function
tool_call
Create a tool call.
Function
tool_call_chunk
Create a tool call chunk.
Function
invalid_tool_call
Create an invalid tool call.
Function
default_tool_parser
Best-effort parsing of tools.
Function
default_tool_chunk_parser
Best-effort parsing of tool chunks.
Function
is_data_content_block
Check if the provided content block is a data content block.
Function
create_text_block
Create a TextContentBlock.
Function
create_image_block
Create an ImageContentBlock.
Function
create_video_block
Create a VideoContentBlock.
Function
create_audio_block
Create an AudioContentBlock.
Function
create_file_block
Create a FileContentBlock.
Function
create_plaintext_block
Create a PlainTextContentBlock.
Function
create_tool_call
Create a ToolCall.
Function
create_reasoning_block
Create a ReasoningContentBlock.
Function
create_citation
Create a Citation.
Function
create_non_standard_block
Create a NonStandardContentBlock.
Function
merge_content
Merge multiple message contents.
Function
message_to_dict
Convert a Message to a dictionary.
Function
messages_to_dict
Convert a sequence of Messages to a list of dictionaries.
Function
get_msg_title_repr
Get a title representation for a message.
Function
get_buffer_string
Convert a sequence of messages to strings and concatenate them into one string.
Function
messages_from_dict
Convert a sequence of messages from dicts to Message objects.
Function
message_chunk_to_message
Convert a message chunk to a Message.
Function
convert_to_messages
Convert a sequence of messages to a list of messages.
Function
filter_messages
Filter messages based on name, type or id.
Function
merge_message_runs
Merge consecutive Messages of the same type.
Function
trim_messages
Trim messages to be below a token count.
Function
convert_to_openai_messages
Convert LangChain messages into OpenAI message dicts.
Function
count_tokens_approximately
Approximate the total number of tokens in messages.
Function
register_translator
Register content translators for a provider in PROVIDER_TRANSLATORS.
Function
get_translator
Get the translator functions for a provider.
Function
translate_content
Derive standard content blocks from a message with Bedrock content.
Function
translate_content_chunk
Derive standard content blocks from a message chunk with Bedrock content.
Function
translate_content
Derive standard content blocks from a message with Anthropic content.
Function
translate_content_chunk
Derive standard content blocks from a message chunk with Anthropic content.
Function
translate_content
Derive standard content blocks from a message with groq content.
Function
translate_content_chunk
Derive standard content blocks from a message chunk with groq content.
Function
translate_content
Derive standard content blocks from a message with Bedrock Converse content.
Function
translate_content_chunk
Derive standard content blocks from a chunk with Bedrock Converse content.
Function
translate_grounding_metadata_to_citations
Translate Google AI grounding metadata to LangChain Citations.
Function
translate_content
Derive standard content blocks from a message with Google (GenAI) content.
Function
translate_content_chunk
Derive standard content blocks from a chunk with Google (GenAI) content.
Function
convert_to_openai_image_block
Convert ImageContentBlock to format expected by OpenAI Chat Completions.
Function
convert_to_openai_data_block
Format standard data content block to format expected by OpenAI.
Function
translate_content
Derive standard content blocks from a message with OpenAI content.
Function
translate_content_chunk
Derive standard content blocks from a message chunk with OpenAI content.
Function
validate_safe_url
Validate a URL for SSRF protection.
Function
is_safe_url
Non-throwing version of validate_safe_url.
Function
validate_resolved_ip
Validate a resolved IP address against the SSRF policy.
Function
validate_hostname
Validate a hostname against the SSRF policy.
Function
validate_url
Validate a URL against the SSRF policy, including DNS resolution.
Function
validate_url_sync
Synchronous URL validation (no DNS resolution).
Function
ssrf_safe_client
Create an httpx.Client with SSRF protection.
Function
ssrf_safe_async_client
Create an httpx.AsyncClient with SSRF protection.
Function
jinja2_formatter
Format a template using jinja2.
Function
validate_jinja2
Validate that the input variables are valid for the template.
Function
mustache_formatter
Format a template using mustache.
Function
mustache_template_vars
Get the top-level variables from a mustache template.
Function
mustache_schema
Get the variables from a mustache template.
Function
validate_f_string_template
Validate an f-string template and return its input variables.
Function
check_valid_template
Check that template string is valid.
Function
get_template_variables
Get the variables from the template.
Function
is_subsequence
Return True if child is subsequence of parent.
Function
format_document
Format a document into a string based on a prompt template.
Function
aformat_document
Async format a document into a string based on a prompt template.
Function
sorted_values
Return a list of values in dict sorted by key.
Function
wait_for_all_evaluators
Wait for all tracers to finish.
Function
run_to_dict
Convert run to dict, compatible with both Pydantic v1 and v2.
Function
run_copy
Copy run, compatible with both Pydantic v1 and v2.
Function
run_construct
Construct run without validation, compatible with both Pydantic v1 and v2.
Function
pydantic_to_dict
Convert any Pydantic model to dict, compatible with both v1 and v2.
Function
pydantic_copy
Copy any Pydantic model, compatible with both v1 and v2.
Function
log_error_once
Log an error once.
Function
wait_for_all_tracers
Wait for all tracers to finish.
Function
get_client
Get the client.
Function
try_json_stringify
Try to stringify an object to JSON.
Function
elapsed
Get the elapsed time of a run.
Function
tracing_v2_enabled
Instruct LangChain to log all runs in context to LangSmith.
Function
collect_runs
Collect all run traces in context.
Function
register_configure_hook
Register a configure hook.
Function
index
Index data from the loader into the vector store.
Function
aindex
Async index data from the loader into the vector store.
Function
tool
Convert Python functions and Runnables to LangChain tools.
Function
convert_runnable_to_tool
Convert a Runnable into a BaseTool.
Function
render_text_description
Render the tool name and description in plain text.
Function
render_text_description_and_args
Render the tool name, description, and args in plain text.
Function
create_retriever_tool
Create a tool to do retrieval of documents.
Function
create_schema_from_function
Create a Pydantic schema from a function's signature.
Function
get_all_basemodel_annotations
Get all annotations from a Pydantic BaseModel and its parents.
Function
default_init_validator
Default init validator that blocks jinja2 templates.
Function
loads
Revive a LangChain class from a JSON string.
Function
load
Revive a LangChain class from a JSON object.
Function
default
Return a default value for an object.
Function
dumps
Return a JSON string representation of an object.
Function
dumpd
Return a dict representation of an object.
Function
try_neq_default
Try to determine if a value is different from the default.
Function
to_json_not_implemented
Serialize a "not implemented" object.
Function
prefix_config_spec
Prefix the id of a ConfigurableFieldSpec.
Function
make_options_spec
Make options spec.
Function
draw_ascii
Build a DAG and draw it in ASCII.
Function
draw_mermaid
Draws a Mermaid graph using the provided graph data.
Function
draw_mermaid_png
Draws a Mermaid graph as PNG using provided syntax.
Function
set_config_context
Set the child Runnable config + tracing context.
Function
ensure_config
Ensure that a config is a dict with all keys present.
Function
get_config_list
Get a list of configs from a single config or a list of configs.
Function
patch_config
Patch a config with new values.
Function
merge_configs
Merge multiple configs into one.
Function
call_func_with_variable_args
Call function that may optionally accept a run_manager and/or config.
Function
acall_func_with_variable_args
Async call function that may optionally accept a run_manager and/or config.
Function
get_callback_manager_for_config
Get a callback manager for a config.
Function
get_async_callback_manager_for_config
Get an async callback manager for a config.
Function
get_executor_for_config
Get an executor for a config.
Function
run_in_executor
Run a function in an executor.
Function
identity
Identity function.
Function
aidentity
Async identity function.
Function
is_uuid
Check if a string is a valid UUID.
Function
node_data_str
Convert the data of a node to a string.
Function
node_data_json
Convert the data of a node to a JSON-serializable format.
Function
coerce_to_runnable
Coerce a Runnable-like object into a Runnable.
Function
chain
Decorate a function to make it a Runnable.
Function
gated_coro
Run a coroutine with a semaphore.
Function
gather_with_concurrency
Gather coroutines with a limit on the number of concurrent coroutines.
Function
accepts_run_manager
Check if a callable accepts a run_manager argument.
Function
accepts_config
Check if a callable accepts a config argument.
Function
accepts_context
Check if a callable accepts a context argument.
Function
asyncio_accepts_context
Check if asyncio.create_task accepts a context arg.
Function
coro_with_context
Await a coroutine with a context.
Function
get_function_first_arg_dict_keys
Get the keys of the first argument of a function if it is a dict.
Function
get_lambda_source
Get the source code of a lambda function.
Function
get_function_nonlocals
Get the nonlocal variables accessed by a function.
Function
indent_lines_after_first
Indent all lines of text after the first line.
Function
add
Add a sequence of addable objects together.
Function
aadd
Asynchronously add a sequence of addable objects together.
Function
get_unique_config_specs
Get the unique config specs from a sequence of config specs.
Function
is_async_generator
Check if a function is an async generator.
Function
is_async_callable
Check if a function is async.
Function
nested_element
Get nested element from path.
Function
droplastn
Drop the last n elements of an iterator.
Function
parse_tool_call
Parse a single tool call.
Function
make_invalid_tool_call
Create an InvalidToolCall from a raw tool call.
Function
parse_tool_calls
Parse a list of tool calls.
Function
py_anext
deprecatedPure-Python implementation of anext() for testing purposes.
Function
get_pydantic_major_version
deprecatedDEPRECATED - Get the major version of Pydantic.
Function
load_prompt_from_config
deprecatedLoad prompt from config dict.
Function
load_prompt
deprecatedUnified method for loading a prompt from LangChainHub or local filesystem.
Modules
Module
langchain_core
langchain-core defines the base abstractions for the LangChain ecosystem.
Module
globals
Global values and configuration that apply to all of LangChain.
Module
rate_limiters
Interface for a rate limiter and an in-memory rate limiter.
Module
sys_info
Print information about the system and langchain packages for debugging purposes.
Module
retrievers
Retriever class returns Document objects given a text query.
Module
stores
Store implements the key-value stores and storage helpers.
Module
exceptions
Custom exceptions for LangChain.
Module
prompt_values
Prompt values for language model prompts.
Module
agents
Schema definitions for representing agent actions, observations, and return values.
Module
cross_encoders
Cross Encoder interface.
Module
version
Version information for langchain-core.
Module
chat_loaders
Chat loaders.
Module
env
Utilities for getting information about the runtime environment.
Module
structured_query
Internal representation of a structured query language.
Module
chat_sessions
Chat Sessions are a collection of messages and function calls.
Module
chat_history
Chat message history stores a history of the message interactions in a chat.
Module
caches
Optional caching layer for language models.
Module
documents
Documents module for data retrieval and processing workflows.
Module
transformers
Document transformers.
Module
compressor
Document compressor.
Module
base
Base classes for media and documents.
Module
language_models
Core language model abstractions.
Module
fake
Fake LLMs for testing purposes.
Module
chat_model_stream
Per-message streaming objects for content-block protocol events.
Module
chat_models
Chat models for conversational AI.
Module
fake_chat_models
Fake chat models for testing purposes.
Module
model_profile
Model profile types and utilities.
Module
llms
Base interface for traditional large language models (LLMs) to expose.
Module
base
Base language models class.
Module
vectorstores
Vector stores.
Module
in_memory
In-memory vector store.
Module
base
A vector store stores embedded data and performs vector search.
Module
utils
Internal utilities for the in memory implementation of VectorStore.
Module
document_loaders
Document loaders.
Module
langsmith
LangSmith document loader.
Module
blob_loaders
Schema for Blobs and Blob Loaders.
Module
base
Abstract interface for document loader implementations.
Module
utils
Utility functions for LangChain.
Module
usage
Usage utilities.
Module
mustache
Adapted from https://github.com/noahmorrison/chevron.
Module
input
Handle chained inputs.
Module
iter
Utilities for working with iterators.
Module
json_schema
Utilities for JSON Schema.
Module
html
Utilities for working with HTML.
Module
uuid
UUID utility functions.
Module
function_calling
Methods for creating function specs in the style of OpenAI Functions.
Module
aiter
Asynchronous iterator utilities.
Module
pydantic
Utilities for pydantic.
Module
json
Utilities for JSON.
Module
strings
String utilities.
Module
env
Utilities for environment variables.
Module
interactive_env
Utilities for working with interactive environments.
Module
formatting
Utilities for formatting strings.
Module
utils
Generic utility functions.
Module
image
Utilities for image processing.
Module
beta_decorator
Helper functions for marking parts of the LangChain API as beta.
Module
deprecation
Helper functions for deprecating parts of the LangChain API.
Module
path
Module
internal
Module
callbacks
Callback handlers allow listening to events in LangChain.
Module
usage
Callback Handler that tracks AIMessage.usage_metadata.
Module
manager
Run managers.
Module
streaming_stdout
Callback Handler streams to stdout on new llm token.
Module
stdout
Callback handler that prints to std out.
Module
file
Callback handler that writes to a file.
Module
base
Base callback handler for LangChain.
Module
outputs
Output classes.
Module
chat_generation
Chat generation output classes.
Module
chat_result
Chat result schema.
Module
generation
Generation output schema.
Module
llm_result
LLMResult class.
Module
run_info
RunInfo class.
Module
messages
Messages are objects used in prompts and chat conversations.
Module
modifier
Message responsible for deleting other messages.
Module
chat
Chat Message.
Module
system
System message.
Module
ai
AI message.
Module
human
Human message.
Module
tool
Messages for tools.
Module
content
Standard, multimodal content blocks for Large Language Model I/O.
Module
function
Function Message.
Module
base
Base message.
Module
utils
Module contains utility functions for working with messages.
Module
block_translators
Derivations of standard content blocks from provider content.
Module
bedrock
Derivations of standard content blocks from Bedrock content.
Module
anthropic
Derivations of standard content blocks from Anthropic content.
Module
langchain_v0
Derivations of standard content blocks from LangChain v0 multimodal content.
Module
groq
Derivations of standard content blocks from Groq content.
Module
bedrock_converse
Derivations of standard content blocks from Amazon (Bedrock Converse) content.
Module
google_genai
Derivations of standard content blocks from Google (GenAI) content.
Module
google_vertexai
Derivations of standard content blocks from Google (VertexAI) content.
Module
openai
Derivations of standard content blocks from OpenAI content.
Module
prompts
A prompt is the input to the model.
Module
dict
Dictionary prompt template.
Module
chat
Chat prompt template.
Module
string
BasePrompt schema definition.
Module
few_shot_with_templates
Prompt template that contains few shot examples.
Module
few_shot
Prompt template that contains few shot examples.
Module
message
Message prompt templates.
Module
prompt
Prompt schema definition.
Module
loading
Load prompts.
Module
structured
Structured prompt template for a language model.
Module
base
Base class for prompt templates.
Module
image
Image prompt template for a multimodal model.
Module
example_selectors
Example selectors.
Module
length_based
Select examples based on length.
Module
semantic_similarity
Example selector that selects examples based on SemanticSimilarity.
Module
base
Interface for selecting examples to include in prompts.
Module
tracers
Tracers are classes for tracing runs.
Module
event_stream
Internal tracer to power the event stream API.
Module
evaluation
A tracer that runs evaluators over completed runs.
Module
run_collector
A tracer that collects all nested runs in a list.
Module
log_stream
Tracer that streams run logs to a stream.
Module
langchain
A tracer implementation that records to LangChain endpoint.
Module
stdout
Tracers that print to the console.
Module
root_listeners
Tracers that call listeners.
Module
memory_stream
Module implements a memory stream for communication between two co-routines.
Module
schemas
Schemas for tracers.
Module
core
Utilities for the root listener.
Module
context
Context management for tracers.
Module
base
Base interfaces for tracing runs.
Module
embeddings
Embeddings.
Module
embeddings
Embeddings interface.
Module
fake
Module contains a few fake embedding models for testing purposes.
Module
indexing
Code to help indexing data into a vectorstore.
Module
api
Module contains logic for indexing documents into vector stores.
Module
in_memory
In memory document index.
Module
base
Base classes for indexing.
Module
tools
Tools are classes that an Agent uses to interact with the world.
Module
convert
Convert functions and runnables to tools.
Module
simple
Tool that takes in function or coroutine directly.
Module
render
Utilities to render tools.
Module
structured
Structured tool.
Module
retriever
Retriever tool.
Module
base
Base classes and utilities for LangChain tools.
Module
load
Load module helps with serialization and deserialization.
Module
load
Load LangChain objects from JSON strings or objects.
Module
mapping
Serialization mapping.
Module
dump
Serialize LangChain objects to JSON.
Module
serializable
Serializable base class.
Module
runnables
LangChain Runnable and the LangChain Expression Language (LCEL).
Module
schema
Module contains typedefs that are used with Runnable objects.
Module
configurable
Runnable objects that can be dynamically configured.
Module
graph_ascii
Draws DAG in ASCII.
Module
graph_png
Helper class to draw a state graph into a PNG file.
Module
graph_mermaid
Mermaid graph drawing utilities.
Module
fallbacks
Runnable that can fallback to other Runnable objects if it fails.
Module
config
Configuration utilities for Runnable objects.
Module
history
Runnable that manages chat message history for another Runnable.
Module
branch
Runnable that selects which branch to run based on a condition.
Module
retry
Runnable that retries a Runnable if it fails.
Module
passthrough
Implementation of the RunnablePassthrough.
Module
graph
Graph used in Runnable objects.
Module
router
Runnable that routes to a set of Runnable objects.
Module
base
Base classes and utilities for Runnable objects.
Module
utils
Utility code for Runnable objects.
Module
output_parsers
OutputParser classes parse the output of an LLM call into structured data.
Module
format_instructions
Format instructions.
Module
xml
Output parser for XML format.
Module
openai_functions
Parsers for OpenAI functions output.
Module
string
String output parser.
Module
pydantic
Output parsers using Pydantic.
Module
list
Parsers for list output.
Module
openai_tools
Parse tools for OpenAI tools output.
Module
json
Parser for JSON output.
Module
base
Base parser for language model outputs.
Module
transform
Base classes for output parsers that can handle streaming input.
Types
Type
PathLike
Type
LanguageModelInput
Input to a language model.
Type
LanguageModelOutput
Output from a language model.
Type
Matrix
Type
Scopes
Type
PydanticBaseModel
Type
TypeBaseModel
Type
Callbacks
Type
Annotation
A union of all defined Annotation types.
Type
DataContentBlock
A union of all defined multimodal data ContentBlock types.
Type
ToolContentBlock
Type
ContentBlock
A union of all defined ContentBlock types and aliases.
Type
MessageLikeRepresentation
A type representing the various ways a message can be represented.
Type
MessageLike
Type
MessageLikeRepresentation
Type
Listener
Type
AsyncListener
Type
ArgsSchema
Type
MessageContentBlock
A single message content block: plain text or a structured block.
Type
ToolExceptionHandlerOutput
Content returned by a handle_tool_error callable.
Type
StreamEvent
Type
MessagesOrDictWithMessages
Type
RunnableLike
Type
AnyConfigurableField
Type