How to Authenticate with Your LLM

LLM Prerequisites

To access LLM models through Eidolon, you may need to:

Create an account with the LLM provider
Fund the account
Acquire an API key

If you receive an authentication error, it may be because Eidolon does not have an API key, or the account is not funded.

Use the table below to get started with popular LLM providers.

LLM	Provider Documentation	API_Key attribute
Claude	Anthropic API Keys, billing and usage	ANTHROPIC_API_KEY
Mistral	Mistral API Keys, billing and usage	MISTRAL_API_KEY
Ollama	Ollama Github	n/a — you run Ollama locally
ChatGPT	OpenAI API Keys, billing and usage	OPENAI_API_KEY
Azure	Azure is different since you need to deploy individual models. See our Azure Agent Example for a more detailed walkthrough.	AZURE_OPENAI_API_KEY

Note that each provider has its own terms and conditions regarding account payment and usage pricing.

👉 Hot tip! Some LLM providers offer evaluation credits. It’s a great way to get started and save money.

Prompts During Installations

If you are prompted to enter an API key during an installation process, it will automatically be added to the Eidolon .env file.

To see your Eidolon environment settings:

cd <Eidolon application root directory>
cat .env

Setting New LLM API Keys

If you want to use a new LLM, you may need to enter an API Key to the .env file. See LLM Prerequisites above to learn about authenticating with various providers.

Example .env file:

OPENAI_API_KEY=youropenaiapikey
ANTHROPIC_API_KEY=youranthropicapikey
MISTRAL_API_KEY=yourmistralapikey

Restart Eidolon. Agents can now authenticate with your preferred LLMs.

Note: if the .env file does not exist in your application root directory, create it as a plain text file.