Environment:Langchain ai Langgraph Python Runtime Environment

Knowledge Sources	LangGraph LangGraph Docs
Domains	Infrastructure, Python
Last Updated	2026-02-11 14:00 GMT

Overview

Python 3.10+ runtime environment with core dependencies for building stateful, multi-actor LLM applications using LangGraph.

Description

This environment defines the base Python runtime and package dependencies required to use the LangGraph framework. All LangGraph libraries require Python 3.10 or higher, with support for CPython and PyPy implementations. The core `langgraph` package depends on `langchain-core` for runnable abstractions, `langgraph-checkpoint` for state persistence, `langgraph-sdk` for API client access, `langgraph-prebuilt` for high-level agent APIs, `pydantic` v2 for data validation, and `xxhash` for fast hashing in cache key generation.

Usage

Use this environment for any LangGraph workflow, whether building a stateful graph, creating a ReAct agent, or deploying via the CLI. This is the mandatory base environment for all LangGraph development.

System Requirements

Category	Requirement	Notes
OS	Linux, macOS, or Windows	All major platforms supported
Python	>= 3.10	CPython or PyPy; 3.10, 3.11, 3.12, 3.13 officially classified
Disk	500MB+	For packages and virtual environment

Dependencies

System Packages

No OS-level system packages required for the base library

Python Packages

`langgraph` >= 1.0.8
`langchain-core` >= 0.1
`langgraph-checkpoint` >= 2.1.0, < 5.0.0
`langgraph-sdk` >= 0.3.0, < 0.4.0
`langgraph-prebuilt` >= 1.0.7, < 1.1.0
`pydantic` >= 2.7.4
`xxhash` >= 3.5.0

Optional Packages (Checkpoint Backends)

`langgraph-checkpoint-sqlite` >= 3.0.3 (requires `aiosqlite` >= 0.20, `sqlite-vec` >= 0.1.6)
`langgraph-checkpoint-postgres` >= 3.0.4 (requires `psycopg` >= 3.2.0, `psycopg-pool` >= 3.2.0, `orjson` >= 3.10.1)
`pycryptodome` (for AES encryption of checkpoint data)

Credentials

The following environment variables may be required depending on usage:

`LANGGRAPH_API_KEY`: Primary API key for connecting to a LangGraph server
`LANGSMITH_API_KEY`: Fallback API key (checked if `LANGGRAPH_API_KEY` is not set)
`LANGCHAIN_API_KEY`: Secondary fallback API key
`LANGGRAPH_AES_KEY`: AES encryption key for encrypted checkpoint serialization (must be 16, 24, or 32 bytes)
`LANGGRAPH_DEFAULT_RECURSION_LIMIT`: Override the default recursion limit (default: 10000)

Quick Install

# Install core LangGraph
pip install langgraph

# With SQLite persistence
pip install langgraph langgraph-checkpoint-sqlite

# With Postgres persistence
pip install langgraph langgraph-checkpoint-postgres

# With encryption support
pip install langgraph pycryptodome

Code Evidence

Python version requirement from `libs/langgraph/pyproject.toml:10`:

requires-python = ">=3.10"

Core dependencies from `libs/langgraph/pyproject.toml:26-33`:

dependencies = [
    "langchain-core>=0.1",
    "langgraph-checkpoint>=2.1.0,<5.0.0",
    "langgraph-sdk>=0.3.0,<0.4.0",
    "langgraph-prebuilt>=1.0.7,<1.1.0",
    "xxhash>=3.5.0",
    "pydantic>=2.7.4",
]

API key resolution from `libs/sdk-py/langgraph_sdk/client.py:89-111`:

def _get_api_key(api_key: str | None = NOT_PROVIDED) -> str | None:
    """Get the API key from the environment.
    Precedence:
        1. explicit string argument
        2. LANGGRAPH_API_KEY (if api_key not provided)
        3. LANGSMITH_API_KEY (if api_key not provided)
        4. LANGCHAIN_API_KEY (if api_key not provided)
    """
    if isinstance(api_key, str):
        return api_key
    if api_key is NOT_PROVIDED:
        for prefix in ["LANGGRAPH", "LANGSMITH", "LANGCHAIN"]:
            if env := os.getenv(f"{prefix}_API_KEY"):
                return env.strip().strip('"').strip("'")
    return None

Encryption key validation from `libs/checkpoint/langgraph/checkpoint/serde/encrypted.py:54-59`:

key_str = os.getenv("LANGGRAPH_AES_KEY")
if key_str is None:
    raise ValueError("LANGGRAPH_AES_KEY environment variable is not set.")
key = key_str.encode()
if len(key) not in (16, 24, 32):
    raise ValueError("LANGGRAPH_AES_KEY must be 16, 24, or 32 bytes long.")

Common Errors

Error Message	Cause	Solution
`ImportError: Pycryptodome is not installed`	Using `EncryptedSerializer` without pycryptodome	`pip install pycryptodome`
`ValueError: LANGGRAPH_AES_KEY environment variable is not set`	Using AES encryption without setting the key	Set `LANGGRAPH_AES_KEY` env var (16, 24, or 32 bytes)
`ValueError: LANGGRAPH_AES_KEY must be 16, 24, or 32 bytes long`	AES key has incorrect length	Use a key of exactly 16, 24, or 32 bytes
`TypeError: Invalid checkpointer provided`	Passing wrong type to `compile(checkpointer=...)`	Use `InMemorySaver`, `SqliteSaver`, `AsyncPostgresSaver`, `True`, `False`, or `None`

Compatibility Notes

Python 3.11+: Required for the CLI in-memory server (`langgraph-cli[inmem]` extra)
Python 3.12+: Some Pydantic internals and asyncio context features are conditionally enabled
Python 3.13+: SDK runtime module has version-specific code paths
Python < 3.14: Some serialization features have this upper bound
Pydantic v2 only: LangGraph requires Pydantic >= 2.7.4; Pydantic v1 is not supported as a primary dependency
typing_extensions: Uses `TypedDict` from `typing_extensions` (not `typing`) for cross-version compatibility

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment