LocalLLMChecker

Unified memory · Apple

What AI models can an Apple M3 8GB run?

8 GB is the awkward size. It runs the 7 to 9 billion parameter class comfortably and hits a wall immediately above it, so most of the decisions here are about context length rather than model choice. Memory is shared with the CPU, so the practical ceiling is lower than the sticker number: macOS hands roughly three quarters of it to the GPU by default.

Memory8 GB
Bandwidth100 GB/s
Usable for a model5.36 GB
Runtime backendMETAL

The short answer

Assuming an 8k context window and default settings, these are the models worth downloading first.

Best all-rounder

Qwen3.5 4B

Q8_0 · 5.24 GB · about 13.8 tokens/s

Best for code

Starcoder2 3B

Q8_0 · 3.46 GB · about 22.9 tokens/s

Largest that still runs well

Qwen3.5 4B

4.66B parameters · Q8_0 · about 13.8 tokens/s

Fastest useful answer

Starcoder2 3B

about 22.9 tokens/s · 3.46 GB

See how fast it feels

Apple M3 8GB running Qwen3.5 4B at Q8_0

Wait for the first word3.9 s
Then writes at13.8 tok/s
Whole answer16 s

YouWhy does my model use more memory when the conversation gets longer?

Model

Simulated from our estimate at a 512-token question, not a recording. Assumes nothing else is competing for the GPU.

Where it sits

We do not track a price for this device, so it does not appear in the budget tools.

Every model, scored on this device

Each row uses the highest-quality quantization that both fits and stays conversational. Speed is a single-stream estimate at 8k context.

Model Params Verdict Download Memory used Speed Max context
Qwen3.5 4B
Qwen 		4.66B Just fits Q8_0 5.24 GB 13.8 t/s 8k
Gemma 3 4B Instruct
Gemma 		4.3B Just fits Q8_0 4.88 GB 14.9 t/s 16k
Qwen2.5 3B Instruct
Qwen 		3.09B Runs well Q8_0 3.65 GB 20.4 t/s 32k
Granite 4.1 3B
Granite 		3.4B Runs well Q8_0 4.31 GB 17 t/s 16k
LFM2.5 2.6B
Liquid 		2.7B Runs well Q8_0 3.11 GB 24.3 t/s 128k
Phi 4 Mini Instruct
Phi 		3.84B Just fits Q8_0 5.14 GB 14.2 t/s 8k
Llama 3.2 3B Instruct
Llama 		3.21B Just fits Q8_0 4.39 GB 16.8 t/s 16k
Qwen3.5 2B
Qwen 		2.27B Runs great Q8_0 2.67 GB 28.8 t/s 128k
Phi 3 Mini 4k Instruct
Phi 		3.82B Just fits Q8_0 4.87 GB 15.3 t/s 4k
Gemma 2 2B Instruct
Gemma 		2.61B Runs well Q8_0 3.51 GB 21.3 t/s 8k
Starcoder2 3B
StarCoder 		3.03B Runs well Q8_0 3.46 GB 22.9 t/s 16k
Qwen3 1.7B
Qwen 		2.03B Runs well Q8_0 3.19 GB 23.6 t/s 16k
DeepSeek R1 Distill Qwen 1.5B
Qwen 		1.78B Runs great Q8_0 2.27 GB 38.9 t/s 64k
Qwen2.5 1.5B Instruct
Qwen 		1.54B Runs great Q8_0 2.03 GB 39 t/s 32k
SmolLM2 1.7B
SmolLM 		1.71B Runs well Q8_0 3.5 GB 21.3 t/s 8k
OLMo 2 0425 1B
OLMo 		1.48B Runs great Q8_0 2.77 GB 30.1 t/s 4k
Llama 3.2 1B Instruct
Llama 		1.24B Runs great Q8_0 1.79 GB 46 t/s 64k
LFM2.5 1.2B Instruct
Liquid 		1.17B Runs great Q8_0 1.56 GB 60.8 t/s 64k
Gemma 3 1B Instruct
Gemma 		1B Runs great Q8_0 1.32 GB 65.4 t/s 32k
Qwen3.5 0.8B
Qwen 		0.87B Runs great Q8_0 1.25 GB 69.9 t/s 256k
Qwen3 0.6B
Qwen 		0.75B Runs great Q8_0 1.89 GB 42 t/s 32k
Qwen2.5 0.5B Instruct
Qwen 		0.49B Runs great Q8_0 0.85 GB 117.5 t/s 32k
LFM2.5 350M
Liquid 		0.35B Runs great Q8_0 0.71 GB 181.9 t/s 64k
DeepSeek Coder V2 Lite Instruct
DeepSeek 		15.71B (2.74B active) Fits only at 2 to 3 bits n/a ! 6.14 GB 32.9 t/s n/a
LFM2.5 8B A1B
Liquid 		8.47B (1.57B active) Just fits n/a 6.87 GB 30.5 t/s n/a
Gemma 4 26B A4B Instruct
Gemma 		25.81B (3.74B active) Fits only at 2 to 3 bits n/a ! 7.34 GB 23 t/s n/a
Gemma 4 E2B Instruct
Gemma 		5.12B Just fits n/a 5.41 GB 13.3 t/s n/a
Qwen3 4B
Qwen 		4.02B Just fits n/a 5.43 GB 13.3 t/s n/a
Internlm3 8B Instruct
InternLM 		8.8B Just fits n/a 6.58 GB 11.6 t/s n/a
Qwen2.5 7B Instruct
Qwen 		7.62B Just fits n/a 6.62 GB 11.6 t/s n/a
Show the other 56 models
Qwen2.5 Coder 7B Instruct 7.62B Just fits n/a 6.62 GB 11.6 t/s n/a
Llama 3.1 8B Instruct 8.03B Just fits n/a 6.7 GB 11.4 t/s n/a
Qwen3.5 9B 9.65B Just fits n/a 7.07 GB 11.3 t/s n/a
Qwen3 8B 8.19B Just fits n/a 6.93 GB 11.1 t/s n/a
Granite 4.1 8B 8.79B Just fits n/a 6.57 GB 11 t/s n/a
Qwen3 14B 14.77B Fits only at 2 to 3 bits n/a 6.92 GB 10.9 t/s n/a
Gemma 4 E4B Instruct 8B Just fits n/a 6.58 GB 10.9 t/s n/a
Gemma 3 12B Instruct 12.19B Just fits n/a 6.73 GB 10.7 t/s n/a
Qwen2.5 VL 7B Instruct 8.29B Just fits n/a 7.13 GB 10.7 t/s n/a
OLMo 3 7B Instruct 7.3B Just fits n/a 6.98 GB 10.7 t/s n/a
DeepSeek Coder 6.7B Instruct 6.74B Fits only at 2 to 3 bits n/a 6.78 GB 10.7 t/s n/a
Mistral 7B Instruct V0.3 7.25B Just fits n/a 6.91 GB 10.6 t/s n/a
Mistral 7B Instruct V0.2 7.24B Just fits n/a 6.9 GB 10.6 t/s n/a
Qwen2.5 14B Instruct 14.77B Fits only at 2 to 3 bits n/a 7.17 GB 10.5 t/s n/a
Qwen2.5 Coder 14B Instruct 14.77B Fits only at 2 to 3 bits n/a 7.17 GB 10.5 t/s n/a
Mistral Nemo Instruct 2407 12.25B Just fits n/a 7.24 GB 10.4 t/s n/a
Gemma 2 9B Instruct 9.24B Just fits n/a 6.9 GB 10.4 t/s n/a
Phi 4 14.66B Fits only at 2 to 3 bits n/a 7.19 GB 10.3 t/s n/a
Mistral Small 24B Instruct 2501 23.57B Fits only at 2 to 3 bits n/a 7.31 GB 10.1 t/s n/a
Gemma 4 12B Instruct 11.96B Just fits n/a 7.13 GB 10.1 t/s n/a
DeepSeek Coder 7B Instruct V1.5 6.91B Just fits n/a 7.27 GB 10.1 t/s n/a
Mixtral 8x7B Instruct V0.1 46.7B CPU only n/a 27.63 GB n/a n/a
Phi 3.5 MoE Instruct 41.87B CPU only n/a 24.92 GB n/a n/a
Qwen3.5 35B A3B 35.95B CPU only n/a 20.74 GB n/a n/a
Qwen3 32B 32.76B CPU only n/a 20.83 GB n/a n/a
Qwen2.5 32B Instruct 32.76B CPU only n/a 20.83 GB n/a n/a
Gemma 4 31B Instruct 31.27B CPU only n/a 21.28 GB n/a n/a
GLM 4.7 Flash 31.22B CPU only n/a 18.28 GB n/a n/a
Qwen3 30B A3B 30.53B CPU only n/a 18.23 GB n/a n/a
Granite 4.1 30B 28.87B CPU only n/a 18.61 GB n/a n/a
Qwen3.5 27B 27.78B CPU only n/a 16.67 GB n/a n/a
Gemma 3 27B Instruct 27.43B CPU only n/a 16.93 GB n/a n/a
Codestral 22B V0.1 22.25B CPU only n/a 14.7 GB n/a n/a
GPT OSS 20B 20.91B CPU only n/a 11.81 GB n/a n/a
Qwen2.5 Coder 32B Instruct 32.76B CPU only n/a 20.83 GB n/a n/a
Qwen3 Coder 30B A3B Instruct 30.53B CPU only n/a 18.23 GB n/a n/a
Kimi K2.6 1026.88B Will not run n/a 578.41 GB n/a n/a
Kimi K2 Instruct 1026.41B Will not run n/a 578.15 GB n/a n/a
GLM 5.1 753.86B Will not run n/a 425.01 GB n/a n/a
GLM 5.2 753.33B Will not run n/a 424.72 GB n/a n/a
DeepSeek V3.2 685.4B Will not run n/a 386.4 GB n/a n/a
DeepSeek R1 684.53B Will not run n/a 385.92 GB n/a n/a
MiniMax M3 427.04B Will not run n/a 241.5 GB n/a n/a
GLM 4.5 358.34B Will not run n/a 204.77 GB n/a n/a
Qwen3 235B A22B 235.09B Will not run n/a 134.03 GB n/a n/a
MiniMax M2.5 228.7B Will not run n/a 130.88 GB n/a n/a
MiniMax M2.7 228.69B Will not run n/a 130.87 GB n/a n/a
Qwen3.5 122B A10B 125.09B Will not run n/a 71.01 GB n/a n/a
GPT OSS 120B 116.83B Will not run n/a 59.65 GB n/a n/a
GLM 4.5 Air 110.47B Will not run n/a 63.93 GB n/a n/a
Qwen3 Next 80B A3B Instruct 81.32B Will not run n/a 46.29 GB n/a n/a
Qwen3 Coder Next 79.67B Will not run n/a 45.37 GB n/a n/a
Qwen2.5 VL 72B Instruct 73.41B Will not run n/a 44.29 GB n/a n/a
Qwen2.5 72B Instruct 72.71B Will not run n/a 43.9 GB n/a n/a
Llama 3.3 70B Instruct 70.55B Will not run n/a 42.68 GB n/a n/a
Llama 3 3 Nemotron Super 49B V1 49.87B Will not run n/a 48.55 GB n/a n/a

What it will not run

63 of the 86 architectures we track are out of reach here, even at 2-bit precision.

If you want more headroom

The next steps up in memory, in order. More memory changes which models load at all; more bandwidth changes how fast they answer, so the two are worth weighing separately.

DeviceMemoryBandwidthModels that fit
GeForce RTX 3080 10GB 10 GB 760 GB/s 53 Check price
Arc B570 10 GB 380 GB/s 53 Check price
GeForce RTX 2080 Ti 11 GB 616 GB/s 57 Check price
GeForce GTX 1080 Ti 11 GB 484 GB/s 57 Check price

Price links go to an Amazon search for the model name and are affiliate links: if you buy through one, we earn a commission at no cost to you. We do not take payment for placement, and the ordering above is by memory capacity alone.

Other devices with 8 GB

Same capacity, different speed. Once a model fits, bandwidth is what separates these.

GeForce RTX 3070 Ti

608 GB/s · NVIDIA

Arc A750

512 GB/s · Intel

GeForce RTX 5060 Ti 8GB

448 GB/s · NVIDIA

GeForce RTX 5060

448 GB/s · NVIDIA

GeForce RTX 3070

448 GB/s · NVIDIA

GeForce RTX 3060 Ti

448 GB/s · NVIDIA

Questions people ask about an Apple M3 8GB

How much VRAM does an Apple M3 8GB have?

The Apple M3 8GB has 8 GB of unified memory shared with the CPU, of which roughly 5.36 GB is available to a model once the operating system takes its share.

Can an Apple M3 8GB run a 70B model?

Not usefully. A 70B model needs about 45.37 GB at 4-bit and this device offers 5.36 GB.

What is the best local model for an Apple M3 8GB?

Qwen3.5 4B at Q8_0 is the strongest all-round fit: 5.24 GB used, around 13.8 tokens per second at 8k context.