GPU GUIDE · APPLE
Best AI models for the
Apple M3 (24GB)
The Apple M3 (24GB) has 24.0GB of unified memory. Below are the top 30 open-source AI models that fit, ranked by composite benchmark score. Each row shows the best quantization that fits your hardware.
VRAM
24.0GB
Brand
apple
Models that fit
30
Generation
M3
Top 30 models for the Apple M3 (24GB)
01
69.0Qwen3.8 27B
27.0BBest fit: Q6_K · 23.2GBCan I run it? →
02
48.3Qwen 3.6 27B
27.8BBest fit: Q6_K · 23.9GBCan I run it? →
03
44.9Qwen3.5-27B
27.0BBest fit: Q6_K · 23.2GBCan I run it? →
04
43.7Qwen 3.6 35B A3B
36.0BBest fit: Q4_K_M · 22.8GBCan I run it? →
05
40.6Muse Glimmer 30B
30.0BBest fit: Q5_K_M · 22.3GBCan I run it? →
06
36.7Devstral 2
24.0BBest fit: Q6_K · 20.8GBCan I run it? →
07
37.0Gemma 4 31B
31.3BBest fit: Q5_K_M · 23.2GBCan I run it? →
08
37.7Qwen 3.5 35B A3B
36.0BBest fit: Q4_K_M · 22.8GBCan I run it? →
09
31.9Gemma 4 26B A4B
26.0BBest fit: Q6_K · 22.4GBCan I run it? →
10
31.7Devstral Small 2
7.0BBest fit: fp16 · 15.0GBCan I run it? →
11
26.7Ministral 3 14B
14.0BBest fit: Q8_0 · 15.9GBCan I run it? →
12
25.9gemma 4 12B it
12.0BBest fit: Q8_0 · 13.8GBCan I run it? →
13
25.0Ministral 3 8B
8.0BBest fit: fp16 · 17.0GBCan I run it? →
14
25.0Gemma 4 E2B
5.1BBest fit: f32 · 21.4GBCan I run it? →
15
24.6Qwen 3.5 9B
9.7BBest fit: fp16 · 20.4GBCan I run it? →
16
23.2Qwen 3.5 4B
4.7BBest fit: f32 · 19.8GBCan I run it? →
17
23.2Granite 4.2 8B
8.0BBest fit: fp16 · 17.0GBCan I run it? →
18
18.6Devstral Small 2 24B Instruct 2512
24.0BBest fit: Q6_K · 20.8GBCan I run it? →
19
18.3Ministral 3 3B
3.0BBest fit: f32 · 13.0GBCan I run it? →
20
15.2GPT-OSS 20B
20.0BBest fit: Q8_0 · 22.3GBCan I run it? →
21
15.0Nemotron 3 Nano Omni (free)
30.0BBest fit: Q5_K_M · 22.3GBCan I run it? →
22
14.3Nemotron 3 Nano 30B A3B
30.0BBest fit: Q5_K_M · 22.3GBCan I run it? →
23
14.3Qwen3 30B A3B Thinking 2507
30.5BBest fit: Q5_K_M · 22.7GBCan I run it? →
24
12.9Qwen3 Coder 30B A3B Instruct
30.5BBest fit: Q5_K_M · 22.7GBCan I run it? →
25
12.7diffusiongemma 26B A4B it
25.8BBest fit: Q6_K · 22.3GBCan I run it? →
26
12.5Qwen3 VL 30B A3B Thinking
31.1BBest fit: Q5_K_M · 23.1GBCan I run it? →
27
12.5QwQ 32B
32.8BBest fit: Q5_K_S · 23.6GBCan I run it? →
28
10.7gemma 4 E4B it
8.0BBest fit: fp16 · 17.0GBCan I run it? →
29
10.3Qwen3 4B Thinking 2507
4.0BBest fit: fp16 · 9.0GBCan I run it? →
30
9.6Qwen3 32B
32.8BBest fit: Q5_K_S · 23.6GBCan I run it? →
Advertisement
FAQ — running AI on the Apple M3 (24GB)
How many AI models can the Apple M3 (24GB) run?
With 24.0GB of unified memory, the Apple M3 (24GB) can run 30+ open-source models from our database, including Qwen3.8 27B, Qwen 3.6 27B, Qwen3.5-27B.
What's the largest LLM I can run on a Apple M3 (24GB)?
The biggest model that fits is approximately 36.0B. Larger models would need to be quantized further or won't fit at all.
Is 24.0GB of unified memory enough for local AI?
Yes — 24.0GB comfortably runs most popular open-source models including 30B-class LLMs at Q4_K_M.