1) GLM is token-efficient compared to DS, Kimi 2) holy shit Zhipu really needs more GPUs. this is sad. *11* million? Not 111? I am pretty sure I'd be getting like an OOM more tokens from $16 worth of DeepSeek than with this monthly plan. …vs V4-Preview, it makes economic sense.
Bought a GLM plan for 3 months, let's see how it goes
Positive users praise GLM Plan as a good model with unbeatable DeepSeek pricing and caching, while others note it is too slow for practical use.
No Digg Deeper questions have been answered for this story yet.
it's a good model
1) GLM is token-efficient compared to DS, Kimi 2) holy shit Zhipu really needs more GPUs. this is sad. *11* million? Not 111? I am pretty sure I'd be getting like an OOM more tokens from $16 worth of DeepSeek than with this monthly plan. …vs V4-Preview, it makes economic sense.
@teortaxesTex only problem, too slow to use it
@teortaxesTex DeepSeek price with caching is unbeatable 🐋