Can internlm2 chat 7b run on the Mac M4 32GB?
Yes. At 8K context the best fit is F16, using about 16.0 GB of the 21.4 GB usable on the Mac M4 32GB. Estimated speed ceiling: ~8 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|
| 2-bit (Q2_K) • | 3.9 GB Fits | 4.4 GB Fits | 7.4 GB Fits | 40 |
| 3-bit (Q3_K_M) • | 4.6 GB Fits | 5.1 GB Fits | 8.1 GB Fits | 31 |
| 4-bit (Q4_K_M) • | 5.5 GB Fits | 6.0 GB Fits | 9.0 GB Fits | 25 |
| 5-bit (Q5_K_M) • | 6.2 GB Fits | 6.7 GB Fits | 9.7 GB Fits | 22 |
| 6-bit (Q6_K) • | 7.0 GB Fits | 7.5 GB Fits | 10.5 GB Fits | 19 |
| 8-bit (Q8_0) • | 8.7 GB Fits | 9.2 GB Fits | 12.2 GB Fits | 15 |
| 16-bit (F16/BF16) | 15.5 GB Fits | 16.0 GB Fits | 19.0 GB Fits | 8 |
• real GGUF file size from QuantFactory/internlm2-chat-7b-GGUF. Speed is a bandwidth ceiling (120 GB/s), not a benchmark. Methodology.
More
internlm2 chat 7b 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 (internlm/internlm2-chat-7b), Apple specs. Data as of 2026-11-17.