Nvidia’s Nemotron 3.5 Super VL gets Fastino training support and Scale enterprise post-training
Scale says its post-trained model matched Skydance’s frontier-model accuracy at 40x lower inference cost.
TLDR
Fastino says its new Training API supports Nvidia’s open-weight Nemotron 3.5 Super VL, while Tinker says the model is live on its platform. Scale says it is working with Nvidia to post-train the model for enterprise workflows. Scale claims its version matched Skydance’s current frontier model’s accuracy at 40x lower inference cost, with responses 5x faster. For EY, Scale says it matched the current frontier model at about one-eighteenth the inference cost, returning answers more than twice as fast.
Combined views
181
1 Source, first seen ago
