Mark Kashef compares OpenAI’s Decisions API with Jev through an image inspection, an outage investigation and a business report check. The image example uses OpenAI alone; the text examples show both systems.
Original by Mark KashefEvaluationIntermediate9 min 38 secPublished
Separate an API’s capabilities from a head-to-head result. The OpenAI-only image demo does not compare image accuracy with Jev.
Compare the evidence, question and choices before comparing answers. The outage and report checks are small demonstrations, not a broad accuracy study.
Read the shown response times and request costs as results for those calls. Different inputs and conditions can change both; the demos do not establish a general winner.
Worth knowing
A small comparison built by the presenter, not an independently repeated benchmark. Official OpenAI documentation supports the API’s input and decision types, but not claims about private model training. No general speed, accuracy or savings claim is adopted.