Skip to content

OpenAI integration

To configure the use of OpenAI models, you need a valid API KEY, which must be available in Brevia as the OPENAI_API_KEY environment variable. One way to set this is by using the BREVIA_ENV_SECRETS configuration in the .env file, as described here.

Embeddings

The basic configuration is as follows:

EMBEDDINGS='{"_type": "openai-embeddings"}'

where openai-embeddings serves as an alias for langchain_openai.embeddings.OpenAIEmbeddings, which can be used interchangeably.

Key variables you can add to this configuration include:

  • model: the name of the OpenAI model to use (string)
  • dimensions: output embeddings size, supported only for text-embedding-3 and later models.

For additional configuration options, refer to the LangChain API reference.

Conversational LLM

Conversational LLMs are configured using the variables QA_COMPLETION_LLM and QA_FOLLOWUP_LLM or SUMMARIZE_LLM for summarization and analysis that use the same JSON format.

An example JSON configuration might look like this:

{"model": "gpt-4o-mini", "model_provider": "openai", "temperature": 0, "max_tokens": 1000}

The primary variables you can add to this configuration include:

  • model: name of the OpenAI model to use (string)
  • temperature: sampling temperature (float, between 0 and 1)
  • max_tokens: maximum number of tokens to generate (int)

See here for more details on the configuration format.

For further configuration options, refer to the LangChain API reference.