1 item tagged with "vg-embed"
Use vg embed to precompute the semantic index so your next vg ask is instant — a local ONNX model downloaded once, then fully offline.