Environment:Langchain ai Langchain OpenAI API Credentials

Knowledge Sources	LangChain OpenAI OpenAI API Reference
Domains	Infrastructure, LLMs
Last Updated	2026-02-11 14:00 GMT

Overview

Environment credentials and SDK dependencies required for OpenAI API integration via `langchain-openai`.

Description

The `langchain-openai` package wraps the official OpenAI Python SDK to provide LangChain-compatible chat models (`ChatOpenAI`), embeddings (`OpenAIEmbeddings`), and structured output. It requires an OpenAI API key and depends on `openai >= 1.109.1` and `tiktoken >= 0.7.0` for token counting. Azure OpenAI is also supported via separate environment variables.

Usage

Required for any workflow that uses OpenAI models including `ChatOpenAI`, `OpenAIEmbeddings`, or Azure OpenAI variants. This includes the Chat_Model_Invocation, Tool_Calling_Structured_Output, Vector_Store_Operations, and Streaming_Responses workflows.

System Requirements

Category	Requirement	Notes
Python	>= 3.10.0, < 4.0.0	Same as core
Network	HTTPS outbound to api.openai.com	Or custom base URL
SDK	openai >= 1.109.1, < 3.0.0	Official OpenAI Python SDK

Dependencies

Python Packages

`langchain-core` >= 1.2.11, < 2.0.0
`openai` >= 1.109.1, < 3.0.0
`tiktoken` >= 0.7.0, < 1.0.0

Credentials

The following environment variables must be set:

Required:

`OPENAI_API_KEY`: OpenAI API key for authentication.

Optional:

`OPENAI_ORG_ID` or `OPENAI_ORGANIZATION`: OpenAI organization ID.
`OPENAI_API_BASE` or `OPENAI_BASE_URL`: Custom API base URL (default: `https://api.openai.com/v1`).

Azure OpenAI (alternative):

`AZURE_OPENAI_API_KEY`: Azure-specific API key.
`AZURE_OPENAI_API_BASE`: Azure endpoint URL.
`AZURE_OPENAI_API_VERSION`: API version string.
`AZURE_OPENAI_DEPLOYMENT_NAME`: Model deployment name.
`AZURE_OPENAI_CHAT_DEPLOYMENT_NAME`: Chat-specific deployment.
`AZURE_OPENAI_EMBEDDINGS_DEPLOYMENT_NAME`: Embeddings-specific deployment.

Quick Install

pip install langchain-openai>=1.1.9

# Set API key
export OPENAI_API_KEY="sk-..."

Code Evidence

Environment variable resolution from `libs/partners/openai/langchain_openai/chat_models/base.py:914-940`:

openai_organization: str | None = Field(
    alias="organization", default=None
)
"""Automatically inferred from env var `OPENAI_ORG_ID` if not provided."""

openai_api_base: str | None = Field(alias="base_url", default=None)
"""Base URL path for API requests, leave blank if not using a proxy or
service emulator."""

Client URL detection from `libs/partners/openai/langchain_openai/chat_models/_client_utils.py:54-66`:

base_url = os.environ.get("OPENAI_BASE_URL", "https://api.openai.com/v1")

Common Errors

Error Message	Cause	Solution
`ValueError: Did not find openai_api_key`	OPENAI_API_KEY not set	`export OPENAI_API_KEY="sk-..."` or pass `api_key=` parameter
`openai.AuthenticationError: 401`	Invalid API key	Verify API key at platform.openai.com
`openai.RateLimitError: 429`	Rate limit exceeded	Implement rate limiting or use `BaseRateLimiter`

Compatibility Notes

OpenAI SDK v1 vs v2: The package supports `openai >= 1.109.1, < 3.0.0`, covering both major versions.
Azure OpenAI: Uses separate environment variables and deployment model. The `AzureChatOpenAI` class handles Azure-specific authentication.
Custom endpoints: Compatible with OpenAI-compatible APIs (e.g., vLLM, Ollama) via `OPENAI_BASE_URL`.

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment