A surface tool result without its images (Media), for the realtime session: a voice model reads a
tool result as text, and gets images as video frames instead. A copy keeps every other key, such as
Success, Data and ErrorMessage. A result without Media, and a value that is not an object,
come back as they are.
A surface tool result without its images (
Media), for the realtime session: a voice model reads a tool result as text, and gets images as video frames instead. A copy keeps every other key, such asSuccess,DataandErrorMessage. A result withoutMedia, and a value that is not an object, come back as they are.