I'm a little confused, because it feels like this doesn't actually talk much about what it was like coding for a month with GLM 5.3 Flash.
The "hurdles" section was about cost and inference provider instability, neither of which has anything to do with whether or not the model was effective at the tasks.
Then we get to "It wasn't all bad", which... you haven't really described anything bad (or at all, really) about the model's performance.
I clicked through to the article because I've started using GLM 5.3 Flash myself (after being exclusively on Opus 5 and prior versions for the past year and a half or so), and I'm curious what other people's experiences have been actually using the model and having it perform tasks. But I feel like I didn't really learn anything from this article.
kelnos · · focus · HN ↗
The "hurdles" section was about cost and inference provider instability, neither of which has anything to do with whether or not the model was effective at the tasks.
Then we get to "It wasn't all bad", which... you haven't really described anything bad (or at all, really) about the model's performance.
I clicked through to the article because I've started using GLM 5.3 Flash myself (after being exclusively on Opus 5 and prior versions for the past year and a half or so), and I'm curious what other people's experiences have been actually using the model and having it perform tasks. But I feel like I didn't really learn anything from this article.