Error messages matching Z.ai GLM I think are the simplest/most compelling. I had Opus 4.8 poke it and it came back with a couple different errors than the article mentions.
Matching the tokenizer is interesting tho
Matching the tokenizer is interesting tho
Comparatively, you can't be 100% sure that Z.ai isn't able to host some other lab's model (although in this case, the hosting errors still support the GLM theory).
You can finetune an LLM to a new tokenizer by nudging just a few layers (even wildly different kinds of tokens), and there's nothing stopping a lab from using someone else's tokenizer.