Skip to content

Can Mac M3 Max 128G run Gemma 4 26B-A4B IT?

Comfortable: GGUF Q8_0 needs 28.5 GB at 4K, batch 1; headroom 67.5 GB.

Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.

Measured runs

No measured run for this pair. Sizing is an estimate, not a runtime test.

Quant memory at 4K

QuantTotal GBVerdictWeight sizing
GGUF Q4_K_M18.2ComfortableVerified file
GGUF Q8_028.5ComfortableVerified file

Context capacity · GGUF Q8_0

ContextTotal GBVerdict
4K28.5Comfortable
16K28.7Comfortable
32K29.1Comfortable
128K31.1Comfortable

Generation speed

GGUF Q8_0: 99 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs