# polling /v1/models every 10s from the moment docker run returned
READY after 160s
{"object":"list","data":[{"id":"google/gemma-4-E2B-it","max_model_len":32768,...}]}