Granite 4.2 8B
Granite 4.2 8B loads in 5 GB at Q4_K_M and is comfortable from 12 GB of memory, Macs included from the Mac Mini M6 16GB. Below: the full memory math, the cheapest card that runs it well, and the fastest machine we track.
PARAMETERS
8.8B
FORMAT
Q4_K_M
MIN MEMORY
8 GB
BEST FOR
Chat, Coding
Memory math
Weights (Q4_K_M)
5 GB
+ KV cache (16k)
~2.0 GB
Total at 16k
~7.0 GB
Comfortable from
12 GB
KV = fp16 estimate (q8_0 cache roughly halves it). "Comfortable" = weights + KV within the engine's tiered budget (~70-85% of memory).
Hardware snapshot
More Granite models
Frequently asked questions
How much memory does Granite 4.2 8B need?
5 GB for the Q4_K_M weights, plus ~2.0 GB of KV-cache at 16k context — about 7.0 GB total. Comfortable from 12 GB of VRAM or unified memory.
Does Granite 4.2 8B run on a Mac?
Yes — from the Mac Mini M6 16GB (~26 tok/s est.). Unified memory means the RAM budget is the only limit.
What is the cheapest GPU for Granite 4.2 8B?
The AMD Radeon RX 7900 XT is the cheapest tracked card that runs Granite 4.2 8B comfortably — ~68 tok/s est. at ~$550 used.
Cite this page
ModelFit: Granite 4.2 8B — specs, memory math and hardware verdicts. https://modelfit.io/models/granite4.2-8b/ (dataset updated 2026-09-03, CC BY 4.0).