Local inference

2 items

Models

IBM shipped a reasoning model in three sizes, and the smallest fits a laptop

Granite 4.2 comes as 3B, 8B and 30B under Apache 2.0, with GGUF and 4-bit builds published alongside the weights. The interesting part is not the benchmark table — it is that the licence and the file formats are settled on day one.

IBMAugust 26verified

Machines & local models

Apple priced the memory, and memory is what a local model actually costs

The new Mac Studio goes to 512GB of unified memory at $5,499; the new Mac mini stops at 32GB and starts at $899. For anyone running models on their own machine, that gap is the whole product line, and Apple is now selling it that way.

AppleAugust 26verified