Use OpenAI with Lumen AI#

Lumen AI ships with the lumen.ai.llm.OpenAI wrapper for OpenAI LLM models. Lumen usually picks between a default and a reasoning model to solve different tasks, by default it will use the following models:

  • default: gpt-4o-mini

  • reasoning: gpt-4o

Prerequisites#

  • Lumen AI installed in your Python environment.

  • An OpenAI API Key. You can obtain this from the OpenAI Dashboard.

Using Environment Variables#

  1. Set the Environment Variable

Set the OPENAI_API_KEY environment variable in your system. This allows Lumen AI to automatically detect and use OpenAI as the provider.

export OPENAI_API_KEY='your-openai-api-key'

set OPENAI_API_KEY=your-openai-api-key

  1. Simply run your Lumen AI application, and it will automatically use OpenAI as the LLM provider. If you have environment variables for multiple providers override it.

lumen-ai serve <your-data-file-or-url> [--provider openai]

Using CLI Arguments#

Alternatively you can also provide the API key as a CLI argument:

lumen-ai serve <your-data-file-or-url> --provider openai --api-key <your-openai-api-key>

Using Python#

In Python, simply import the LLM wrapper lumen.ai.llm.OpenAI and pass it to the lumen.ai.ui.ExplorerUI:

import lumen.ai as lmai

openai_llm = lmai.llm.OpenAI('<your-data-file-or-url>', api_key='your-openai-api-key')

ui = lmai.ui.ExplorerUI(llm=openai_llm)
ui.servable()

Configuring the Model#

If you want to override the default and/or reasoning models, which use GPT-4o-mini and GPT-4o respectively, a model configuration can be provided via the model_kwargs parameter. Lumen will use different models for different scenarios, currently a default and a reasoning model can be provided. If no reasoning model is provided it will always use the default model.

You can find the latest list of OpenAI models in their documentation. To swap GPT-4o for o1-mini as the reasoning model simply override the default model_kwargs:

import lumen.ai as lmai

config = {
    "default": "gpt-4o-mini",
    "reasoning": "o1-mini"
}

llm = lmai.llm.OpenAI(model_kwargs=config)

lmai.ui.ExplorerUI('<your-data-file-or-url>', llm=llm).servable()