The best 2-bit quantization on paper loses 2.27x to a trellis kernel once you serve it.

Leech-lattice vector quantization holds the strongest reported 2-bit quality under its own protocol, but no implementation of the multi-shell decoder existed. This paper supplies one for the full 301-class codebook with a fused dequantize-plus-matvec kernel and measures batch-1 decode GEMV cost. The distinction that matters is that in-VRAM rate is a separate axis from on-disk rate: four bit-exact layouts show binary bit planes beating one-hot masks at 4.80 bits per weight. Run in-process against deployed AWQ (4-bit) and QTIP (2-bit) kernels, the trellis kernel reads 2.40x fewer bytes and runs 2.27x faster, with the time gap tracking the traffic gap.