RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can LFM2.5 8B A1B run on the Mac M4 32GB?

Yes. At 8K context the best fit is Q8_0, using about 9.1 GB of the 21.4 GB usable on the Mac M4 32GB. Estimated speed ceiling: ~113 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K)3.6 GB Fits3.7 GB Fits4.0 GB Fits320
3-bit (Q3_K_M)4.5 GB Fits4.6 GB Fits4.9 GB Fits246
4-bit (Q4_K_M) •5.5 GB Fits5.5 GB Fits5.8 GB Fits197
5-bit (Q5_K_M) •6.3 GB Fits6.4 GB Fits6.6 GB Fits169
6-bit (Q6_K) •7.2 GB Fits7.2 GB Fits7.5 GB Fits146
8-bit (Q8_0) •9.1 GB Fits9.1 GB Fits9.4 GB Fits113
16-bit (F16/BF16) •32.3 GB No32.3 GB No32.6 GB No—

• real GGUF file size from LiquidAI/LFM2.5-8B-A1B-GGUF. Speed is a bandwidth ceiling (120 GB/s), not a benchmark. Methodology.

More

LFM2.5 8B A1B on every GPU and Mac · Every model that runs on the Mac M4 32GB · Apple Silicon guide · Try other settings in the calculator

Data: Hugging Face Hub (LiquidAI/LFM2.5-8B-A1B), Apple specs. Data as of 2026-11-17.