Umm, are these guys sending requests at a public inference endpoint or are they doing more sophisticated work like training etc?

Lock in is pretty easy these days. Just a dummy example, Claude models are trained on their `str_replace_based_edit_tool` edit tool[1] which is very different from OpenAI's `apply_patch` tool[2].

1. https://platform.claude.com/docs/en/agents-and-tools/tool-us...

2. https://developers.openai.com/api/docs/guides/tools-apply-pa...