Verathos Beats OpenRouter’s GLM-5.2 Rates by 27%

Verathos (SN96) is now serving GLM-5.2 at $0.25 per million input tokens and $1.00 per million output tokens.

The cheapest equivalent rates on OpenRouter today sit at $0.343 input and $1.078 output per million tokens. That means Verathos (SN96) undercuts the listed alternative on OpenRouter by roughly 27% on the same model.

Verathos (SN96) is a decentralized compute network through which every inference and training operation gets cryptographically proven correct and anchored on-chain.

The lower pricing comes from the mesh infrastructure powering the serving layer underneath, called Sleipnir.

Sleipnir lets multiple operators pool their GPU capacity to serve larger models across distributed consumer hardware.

The pooled compute appears as one endpoint, so a large model runs across many contributors at once. Operators earn from serving excess capacity while using the same pool for their own inference workloads.

The team noted that GLM-5.3 is coming to Verathos soon, expanding access to more open-source frontier models.

➛ Read More on Sleipnir Here, and Access Verathos’ Listed Models Here.