Either using the single model Validation API or the Fluent API for multiple models (yet validating only a single model) should result in the same KIE Message being reported. This is linked to Workbench behavior which relies on Kie Message level and not typed-message for DT Analysis