M2 24GB
Apple Silicon
24GB VRAM · 100 GB/s bandwidth · 7.2 FP16 TFLOPS · 15W TDP · $799 street price
The M2 24GB has 24GB of VRAM. With only 24GB available, no current reference LLMs fit comfortably at Q4 quantization. Consider GPUs with 12GB+ VRAM for local inference.
What LLMs can you run?
| Model | Params | Q4 Weight | Fit | Decode |
|---|---|---|---|---|
| Gemma 3 4B | 4.0B | 2 GB | tight | 21 tok/s |
| Qwen 2.5 7B | 7.6B | 4 GB | tight | 11 tok/s |
| Llama 3.1 8B | 8.0B | 4 GB | tight | 10 tok/s |
| Mistral Small 24B | 24.0B | 12 GB | tight | 3 tok/s |
| Gemma 3 27B | 27.4B | 14 GB | tight | 3 tok/s |
| Qwen 2.5 Coder 32B | 32.5B | 16 GB | tight | 2 tok/s |
| Llama 3.3 70B | 70.6B | 35 GB | won't fit | |
| Qwen 2.5 72B | 72.7B | 36 GB | won't fit | |
| Llama 3.1 405B | 405B | 202 GB | won't fit | |
| DeepSeek R1 671B | 671B | 336 GB | won't fit |
Similar GPUs
| GPU | VRAM | BW | TFLOPS | TDP |
|---|---|---|---|---|
| M3 24GB | 24GB | 100 GB/s | 8.2 | 15W |
| M4 24GB | 24GB | 120 GB/s | 9.2 | 20W |
| M4 Pro 24GB | 24GB | 273 GB/s | 18.4 | 45W |
| Tesla P40 | 24GB | 347 GB/s | 0.2 | 250W |
| RTX 4500 Ada Generation | 24GB | 432 GB/s | 39.6 | 210W |
Compare with another GPU
Select another GPU to compare specs and model performance side by side.