Can LFM2.5 VL 3B run on the Mac M4 Pro 64GB?
Yes. At 8K context the best fit is F16, using about 10.8 GB of the 48.0 GB usable on the Mac M4 Pro 64GB. Estimated speed ceiling: ~25 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|
| 2-bit (Q2_K) | 1.8 GB Fits | 1.9 GB Fits | 2.2 GB Fits | 233 |
| 3-bit (Q3_K_M) | 2.1 GB Fits | 2.2 GB Fits | 2.6 GB Fits | 179 |
| 4-bit (Q4_K_M) • | 2.3 GB Fits | 2.3 GB Fits | 2.7 GB Fits | 163 |
| 5-bit (Q5_K_M) • | 2.5 GB Fits | 2.6 GB Fits | 3.0 GB Fits | 141 |
| 6-bit (Q6_K) • | 2.8 GB Fits | 2.8 GB Fits | 3.2 GB Fits | 123 |
| 8-bit (Q8_0) • | 3.4 GB Fits | 3.4 GB Fits | 3.8 GB Fits | 95 |
| 16-bit (F16/BF16) • | 10.8 GB Fits | 10.8 GB Fits | 11.2 GB Fits | 25 |
• real GGUF file size from LiquidAI/LFM2.5-VL-3B-GGUF. Speed is a bandwidth ceiling (273 GB/s), not a benchmark. Methodology.
More
LFM2.5 VL 3B on every GPU and Mac · Every model that runs on the Mac M4 Pro 64GB · Apple Silicon guide · Try other settings in the calculator
Data: Hugging Face Hub (LiquidAI/LFM2.5-VL-3B), Apple specs. Model data last checked 2026-11-17; refreshed hourly.