Can SmolLM2 360M Instruct run on the Mac M3 24GB?
Yes. At 8K context the best fit is F16, using about 1.5 GB of the 16.1 GB usable on the Mac M3 24GB. Estimated speed ceiling: ~138 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | Max tok/s (est., 8K) |
|---|---|---|---|
| 2-bit (Q2_K) • | 0.9 GB Fits | 1.0 GB Fits | 457 |
| 3-bit (Q3_K_M) • | 0.9 GB Fits | 1.0 GB Fits | 426 |
| 4-bit (Q4_K_M) • | 0.9 GB Fits | 1.1 GB Fits | 370 |
| 5-bit (Q5_K_M) • | 0.9 GB Fits | 1.1 GB Fits | 345 |
| 6-bit (Q6_K) • | 1.0 GB Fits | 1.1 GB Fits | 272 |
| 8-bit (Q8_0) • | 1.0 GB Fits | 1.2 GB Fits | 259 |
| 16-bit (F16/BF16) • | 1.3 GB Fits | 1.5 GB Fits | 138 |
• real GGUF file size from unsloth/SmolLM2-360M-Instruct-GGUF. Speed is a bandwidth ceiling (100 GB/s), not a benchmark. Methodology.
More
SmolLM2 360M Instruct on every GPU and Mac · Every model that runs on the Mac M3 24GB · Apple Silicon guide · Try other settings in the calculator
Data: Hugging Face Hub (HuggingFaceTB/SmolLM2-360M-Instruct), Apple specs. Data as of 2026-11-17.