Inkling
Inkling is a multimodal MoE model (975B total, 41B active, 256k context) reasoning over text, image, and audio inputs.
Select model
Use the selector above to choose a model.