Environment:Run llama Llama index OpenAI API Configuration

Knowledge Sources	LlamaIndex Finetuning OpenAI API Reference
Domains	Infrastructure, LLM_Finetuning, Evaluation
Last Updated	2026-02-11 19:00 GMT

Overview

OpenAI API key configuration required for LLM finetuning, evaluation, and default LLM/embedding model usage in LlamaIndex.

Description

This environment defines the API credentials needed to interact with OpenAI services. The `OPENAI_API_KEY` is used by the `OpenAIFinetuneEngine` for launching and monitoring finetuning jobs, by the evaluation pipeline for LLM-as-judge evaluators, and as the default LLM provider when using the `llama-index` meta-package (which includes `llama-index-llms-openai`).

Usage

Use this environment when running any workflow that calls OpenAI APIs: OpenAI LLM Finetuning, Evaluation Pipeline (with OpenAI evaluator LLM), or RAG Query Pipeline with default OpenAI LLM/embedding configuration.

System Requirements

Category	Requirement	Notes
Network	Internet access to api.openai.com	Required for all API calls
Account	OpenAI API account with billing	Finetuning requires paid tier

Dependencies

Python Packages

`openai` (pulled in by `llama-index-llms-openai`)
`llama-index-llms-openai` >= 0.6.0, < 0.7

Credentials

The following environment variables must be configured:

`OPENAI_API_KEY`: OpenAI API key for authentication. Used by `OpenAIFinetuneEngine` and default LLM/embedding resolution.

For Azure OpenAI users, additional variables are needed:

`AZURE_OPENAI_API_KEY`: Azure-specific API key
`AZURE_OPENAI_ENDPOINT`: Azure endpoint URL
`OPENAI_API_VERSION`: API version (defaults to `2024-02-01`)

Quick Install

# Install OpenAI LLM integration
pip install llama-index-llms-openai>=0.6.0

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

Code Evidence

OpenAI client initialization from `openai/base.py:38`:

self._client = SyncOpenAI(api_key=os.getenv("OPENAI_API_KEY", None))

Azure OpenAI client initialization from `azure_openai/base.py:37-39`:

azure_endpoint=os.getenv("AZURE_OPENAI_ENDPOINT"),
api_key=os.getenv("AZURE_OPENAI_API_KEY", None),
api_version=os.getenv("OPENAI_API_VERSION", "2024-02-01"),

Generic environment resolution pattern from `generic_utils.py:311-329`:

def get_from_param_or_env(
    key: str,
    param: Optional[str] = None,
    env_key: Optional[str] = None,
    default: Optional[str] = None,
) -> str:
    if param is not None:
        return param
    elif env_key and env_key in os.environ and os.environ[env_key]:
        return os.environ[env_key]
    elif default is not None:
        return default
    else:
        raise ValueError(
            f"Did not find {key}, please add an environment variable"
            f" `{env_key}` which contains it, or pass"
            f"  `{key}` as a named parameter."
        )

Common Errors

Error Message	Cause	Solution
`AuthenticationError: Incorrect API key provided`	Invalid or expired OPENAI_API_KEY	Verify API key at platform.openai.com and re-export
`ValueError: Did not find api_key, please add an environment variable`	OPENAI_API_KEY not set	`export OPENAI_API_KEY="sk-..."`
`RateLimitError`	API rate limits exceeded	Reduce batch eval workers or add retry logic
`InsufficientQuotaError`	Billing limits reached	Check OpenAI billing dashboard

Compatibility Notes

Azure OpenAI: Fully supported via `AzureOpenAIFinetuneEngine`. Requires separate credential variables (`AZURE_OPENAI_*`).
API Key Precedence: Constructor parameter > environment variable > raises ValueError. This pattern is consistent across all LlamaIndex LLM integrations.
Finetuning Access: OpenAI finetuning requires a paid account tier. Free-tier accounts cannot create finetuning jobs.

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment