For testing out a few inference providers for #GLM5.2, I created a new #Pi extension for Euro provider #TensorX. They have a smol selection of models (tensorx.ai/models/), but what they serve, they serve *fast*. 200+ tps, constantly. Nice! github.com/czottmann/pi... #LLM
github.comGitHub - czottmann/pi-tensorx: TensorX provider extension for pi coding agent. (TensorX is a European LLM router. I'm a customer.)TensorX provider extension for pi coding agent. (TensorX is a European LLM router. I'm a customer.) - czottmann/pi-tensorx
1 repost1 like