Skip to content

Can Mac M5 Ultra 512G run Ling-3.1-flash?

Tight: GGUF Q4_K_M needs 370.6 GB at 4K, batch 1; headroom 13.4 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_M370.6TightVerified file
GGUF Q8_0627.6Does not fitVerified file

Context capacity · GGUF Q4_K_M

ContextTotal GBVerdict
4K370.6Tight
16K370.7Tight
32K370.9Tight
128K371.9Tight

Generation speed

GGUF Q4_K_M: 80 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs