Motif-2-12.7B-Reasoning
Open weightsOpen weightsText12.7B parameters
Motif family model by Motif Technologies.
About the Motif family
Motif Technologies' Korean/English models (Motif-2 12.7B with Grouped Differential Attention; Motif 3 in beta) trained on trillions of tokens.
All Motif modelsHow this one is configured
This configuration ~12.7B parameters.
Who makes it
Motif Technologies — South Korea. Not listed as a provider in this directory.