How do I use OpenAI decision models?
Add an OpenAI connection with access to gpt-6-luna in a decision-model evaluator. Langfuse calls the OpenAI Decisions API through this connection.
Map observation data to the single state named input. Questions evaluate that state as a whole and cannot reference individual fields. Score levels need a label and can include an optional description.
If OpenAI refuses any question, the run fails and writes no scores.
Was this page helpful?
Last updated on