Hardware · Apple Silicon

What AI models can the Apple M3 (24 GB) run?

Unified memory: 24.0 GB
Usable for models: 16.1 GB
Memory bandwidth: 100 GB/s
FP16 compute: 4.1 TFLOPS
Released: 2023

macOS lets the GPU use roughly two-thirds to three-quarters of unified memory by default, so we plan around that figure rather than the full 24.0 GB.

Specifications

Maker	Apple Silicon
Type	Apple silicon (system on a chip)
Chip	Apple M3
Unified memory	24.0 GB
Usable for AI models	16.1 GB
Memory bandwidth	100 GB/s
FP16 compute	4.1 TFLOPS
Interface	UNIFIED
Released	2023

4 products with the Apple M3 (24 GB)

MacBook Air (13-inch, M3, 2024)Apple · Laptop · 8, 16 or 24 GB · from $1,099Maker’s page (opens in a new tab)
MacBook Air (15-inch, M3, 2024)Apple · Laptop · 8, 16 or 24 GB · from $1,299Maker’s page (opens in a new tab)
MacBook Pro (14-inch, M3, Nov 2023)Apple · Laptop · 8, 16 or 24 GB · from $1,599Maker’s page (opens in a new tab)
iMac (24-inch, 2023) (M3)Apple · Desktop · 8, 16 or 24 GB · from $1,299Maker’s page (opens in a new tab)

Best LLMs for the Apple M3 (24 GB)

Grade · points	Model	Quant	Memory	Speed	Fit
84 ptsGreat (A) · 84 of 100 points: Model quality 35.0/45, Memory fit 24.0/25, Speed 10.1/15, Memory use 15.0/15	Gemma 4 26B-A4BGoogle	`Q3_K_M`	14.9 GB	18.0 tok/s	Tight fit
82 ptsGreat (A) · 82 of 100 points: Model quality 37.0/45, Memory fit 25.0/25, Speed 5.9/15, Memory use 15.0/15	Gemma 4 12BGoogle	`Q6_K`	12.4 GB	6.1 tok/s	Runs well
80 ptsGreat (A) · 80 of 100 points: Model quality 36.1/45, Memory fit 24.0/25, Speed 5.3/15, Memory use 15.0/15	Qwen3.8 27BAlibaba	`Q3_K_M`	16.0 GB	4.9 tok/s	Tight fit
80 ptsGreat (A) · 80 of 100 points: Model quality 34.3/45, Memory fit 25.0/25, Speed 6.5/15, Memory use 14.4/15	Ornith 1.5 9BDeepReinforce	`Q6_K`	10.1 GB	7.6 tok/s	Runs well
79 ptsGreat (A) · 79 of 100 points: Model quality 31.5/45, Memory fit 24.0/25, Speed 9.4/15, Memory use 15.0/15	gpt-oss 20BOpenAI	`MXFP4`	14.0 GB	15.7 tok/s	Tight fit
79 ptsGreat (A) · 79 of 100 points: Model quality 33.9/45, Memory fit 25.0/25, Speed 6.5/15, Memory use 14.4/15	Qwen3.5 9BAlibaba	`Q6_K`	10.1 GB	7.6 tok/s	Runs well
79 ptsGreat (A) · 79 of 100 points: Model quality 35.5/45, Memory fit 24.0/25, Speed 5.3/15, Memory use 15.0/15	Qwen3.6 27BAlibaba	`Q3_K_M`	16.0 GB	4.9 tok/s	Tight fit
79 ptsGreat (A) · 79 of 100 points: Model quality 27.7/45, Memory fit 24.0/25, Speed 12.8/15, Memory use 15.0/15	GLM-4.7 Flash (30B-A3B)Zhipu AI	`Q2_K`	14.8 GB	28.9 tok/s	Tight fit
79 ptsGreat (A) · 79 of 100 points: Model quality 35.2/45, Memory fit 24.0/25, Speed 5.3/15, Memory use 15.0/15	Qwen3.5 27BAlibaba	`Q3_K_M`	16.0 GB	4.9 tok/s	Tight fit
79 ptsGreat (A) · 79 of 100 points: Model quality 33.4/45, Memory fit 25.0/25, Speed 6.5/15, Memory use 14.4/15	Ornith 1.0 9BDeepReinforce	`Q6_K`	10.1 GB	7.6 tok/s	Runs well
79 ptsGreat (A) · 79 of 100 points: Model quality 27.4/45, Memory fit 24.0/25, Speed 12.8/15, Memory use 15.0/15	Qwen AgentWorld 35B-A3BAlibaba	`Q2_K`	15.9 GB	28.9 tok/s	Tight fit
78 ptsGreat (A) · 78 of 100 points: Model quality 29.9/45, Memory fit 24.0/25, Speed 10.0/15, Memory use 15.0/15	ERNIE 4.5 21B-A3B ThinkingBaidu	`Q4_K_M`	15.7 GB	17.6 tok/s	Tight fit

Best Image models for the Apple M3 (24 GB)

Grade · points	Model	Precision	Memory	Time per output	Fit
85 ptsExcellent (S) · 85 of 100 points: Model quality 34.4/45, Memory fit 25.0/25, Speed 11.8/15, Memory use 14.6/15	FLUX.2 [klein] 4BBlack Forest Labs	`FP8`	10.2 GB	16 s	Runs well
84 ptsGreat (A) · 84 of 100 points: Model quality 37.0/45, Memory fit 25.0/25, Speed 7.4/15, Memory use 15.0/15	Z-Image TurboAlibaba (Tongyi)	`FP8`	12.6 GB	41 s	Runs well
80 ptsGreat (A) · 80 of 100 points: Model quality 37.0/45, Memory fit 20.0/25, Speed 8.6/15, Memory use 15.0/15	FLUX.2 [klein] 9BBlack Forest Labs	`FP8`	11.7 GB	31 s	Runs well
79 ptsGreat (A) · 79 of 100 points: Model quality 36.5/45, Memory fit 25.0/25, Speed 2.5/15, Memory use 15.0/15	Z-ImageAlibaba (Tongyi)	`Q8 GGUF`	13.0 GB	5.9 min	Runs well
75 ptsGreat (A) · 75 of 100 points: Model quality 34.3/45, Memory fit 23.0/25, Speed 2.9/15, Memory use 15.0/15	FLUX.1 Krea [dev]Black Forest Labs	`Q4 GGUF`	14.8 GB	4.3 min	Tight fit
75 ptsGreat (A) · 75 of 100 points: Model quality 32.6/45, Memory fit 25.0/25, Speed 2.4/15, Memory use 15.0/15	Chroma1-HDLodestone Rock	`NF4`	12.5 GB	6.6 min	Runs well

Best Video models for the Apple M3 (24 GB)

Grade · points	Model	Precision	Memory	Time per output	Fit
72 ptsGreat (A) · 72 of 100 points: Model quality 22.1/45, Memory fit 25.0/25, Speed 10.0/15, Memory use 15.0/15	LTX-Video 2BLightricks	`FP8`	10.8 GB	6.8 min	Runs well
67 ptsGood (B) · 67 of 100 points: Model quality 24.3/45, Memory fit 25.0/25, Speed 3.2/15, Memory use 15.0/15	Wan 2.1 T2V 1.3BAlibaba	`FP8`	12.8 GB	65.0 min	Runs well
66 ptsGood (B) · 66 of 100 points: Model quality 19.8/45, Memory fit 25.0/25, Speed 6.4/15, Memory use 15.0/15	CogVideoX 2BZhipu AI	`FP8`	11.0 GB	16.3 min	Runs well
65 ptsGood (B) · 65 of 100 points: Model quality 24.3/45, Memory fit 23.0/25, Speed 3.7/15, Memory use 15.0/15	CogVideoX 5BZhipu AI	`FP8`	14.3 GB	48.8 min	Tight fit
65 ptsGood (B) · 65 of 100 points: Model quality 24.3/45, Memory fit 23.0/25, Speed 3.7/15, Memory use 15.0/15	CogVideoX 5B I2VZhipu AI	`FP8`	14.9 GB	48.8 min	Tight fit

Best LLMs by task on this hardware Open in Model Finder

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.