How does the GPT-J inference API work?

GPT-J is supported by the inference API. You can try it out here: EleutherAI/gpt-j-6B · Hugging Face

Can you verify whether you are able to generate more than a given word?