langchain
LangChain library is the foundational component that allows developers to work with LLMs and integrate them with external tools, APIs, and databases.
LangChain integration providers extend the library's capabilities, providing other systems and tools.
LangGraph library: build upon LangChain and provide a more structured way to design, visualize, and manage complex workflows.
LangGraph CLI allows developers to interact with LangGraph in a more direct way, simplifying certain tasks like deployment, testing, and execution.
LangGraph examples were created to help developers understand how to use LangGraph in real-world scenarios.
LangGraph Platform is an integrated environment for managing LangGraph applications, enabling debugging, visualizing, and the monitoring of workflows.
LangSmith Platform is full suite of tools for managing LLM-based applications, debugging, collaboration, and monitoring at a higher level.
LangGraph use-cases describe the functionality of real-world applications that use LangChain and LangGraph.
Also Consider
In addition to the components listed above, there are several other important parts of the LangChain ecosystem that might be missing from the above list. These components help extend LangChain's functionality and make it easier to build and deploy language model-driven applications:
LangChain Agents
Agents are components in LangChain that allow LLMs to perform tasks with dynamic decision-making. They use tools and chains to interact with external resources (e.g., APIs, databases) in a more intelligent, autonomous manner.
LangChain Tools
Tools are external components that LangChain can integrate with, such as APIs, databases, or file systems. LangChain can interact with these tools during the execution of chains or agents to enhance LLM-based workflows.
LangChain Chains
Chains are sequences of steps where the output of one step is fed into the next. LangChain chains are used to build structured workflows, where each step involves either an LLM call or some form of computation.
LangChain Memory
Memory components allow LangChain applications to store and retrieve information across multiple interactions. This is crucial for building conversational agents that remember past interactions.
LangChain Prompts
LangChain includes tools for managing and optimizing prompts, which are essential for generating meaningful interactions with language models. It allows developers to structure, modify, and reuse prompts efficiently.
LangChain Output Parsers
Output parsers process the results returned by language models, structuring them into formats that are usable by the application, such as structured data, JSON, or other formats.
LangChain Document Loaders
These components allow LangChain to ingest documents and data from a variety of sources (e.g., PDFs, web pages, databases) into the system, so they can be processed by language models.
LangChain Embeddings
LangChain provides tools for working with embeddings, which are vector representations of text used for similarity matching, information retrieval, and other tasks.
LangChain Memory Stores
These are storage systems used in LangChain to persist memory, especially in long-running applications like chatbots or virtual assistants that require long-term memory.
LangChain Hub
The LangChain Hub is a repository of pre-built chains, agents, and components. It's a place where developers can share reusable pieces of LangChain code and find components for their own projects.
LangChain Integration with External Services
LangChain integrates with several external services and platforms such as OpenAI, Google Cloud, AWS, Azure, and more. These integrations help expand LangChain’s capabilities by leveraging third-party APIs, models, and tools.
LangChain Evaluation Framework
This framework helps developers assess the performance of their LangChain applications by providing metrics and benchmarks for evaluating the effectiveness of chains, agents, and other components.