upvote
That's my point too.

If you take GLM and call it ChatGPT or Claude Opus is anyone going to notice? If you are not into agentic AI I would argue that the model type makes zero difference for day to day use because GLM 5.2 is hitting the benchmarks hard.

Now for a specialised use case (narrow fields), say cyber, Mythos is possibly better.

reply