3 ms·
>token use is higher than k3 and far higher than proprietary models GLM sets effort to max by default historically.
by Havoc 2mo ago
>token use is higher than k3 and far higher than proprietary models
GLM sets effort to max by default historically.
- markasoftware 2mo agoAa also benchmarked k3 at max