OpenAI CEO Sam Altman introduced the Decisions API during the company's Dev Day event. This tool allows the Luna model to select from a predefined set of options, such as image categories or specific agent behaviors.
The new API functions similarly to Jev, a software automation model released by TypeSafe AI earlier this month. Jev operates as a classifier built on a large language model, providing probabilistic outputs at high speeds and low costs.
TypeSafe CEO Diogo Almeida noted that the industry is shifting toward System One compatible architectures, which prioritize fast and intuitive processing over slower, deliberate reasoning. TypeSafe utilizes synthetic data to ensure the statistical accuracy of its model's outputs.
Security experts suggest that decision models could significantly reduce the costs of monitoring AI agents. Cybersecurity professional Shapor Naghibzadeh demonstrated that using Jev to verify agent actions costs $2.94, compared to $372 when using a standard frontier large language model.
OpenAI has released the Decisions API as a limited preview. While the exact technical similarities between the two products remain unconfirmed, the emergence of such tools reflects a broader trend among tech companies to develop faster, more efficient alternatives to traditional large language models.
