Qwen3.8-27B (bf16) is now registered on TensorCash mainnet, a model reported as competitive with today’s top models.
From now on, coding and autonomous agent tasks can be bought as inference from the TensorCash network, a large step up from Qwen3-8B.
Block producers and miners are advised to upgrade to 1.2.3a-r2 before mining with the new model.
Believing in TensorCash means believing that the outputs of AI models must be verifiable at scale.
Inference is slowly becoming a commodity, and that brings risk. When silver became common trade, it was mixed with cheaper metals, until assay and hallmarks became the standard.
An AI answer is even harder to verify. Unlike silver, it cannot be assayed. Looking at it tells you nothing about which model produced it.
TensorCash makes inference verification a core part of the protocol. No one has to wonder whether they were answered by the model they paid for.
In a world where AI is a commodity, we believe inference that cannot be verified is worth close to zero.