Rent a GPU

Can the Apple M3 Ultra (256GB) run Qwen2.5 Coder 32B?

Apple M3 Ultra (256GB) × Qwen2.5 Coder 32B8K context
YesRuns fully on the GPU
Memory at FP1668.3 GB of 192 GB
Weights 65.6 GBKV cache 2.1 GBRuntime 0.6 GB
QuantFP16Full quality
Speed~7tokens per second
Max context32Kthe model's full window
  1. FP16
  2. Q8_0
  3. Q6_K
  4. Q5_K_M
  5. Q4_K_M
  6. Q3_K_M
  7. Q2_K

Yes. The Apple M3 Ultra (256GB) can run Qwen2.5 Coder 32B fully in unified memory at FP16 (full quality), using about 68.3 GB of the 192 GB available, at roughly 7 tokens/second.

Try other context lengths

Run it with Ollama

With Ollama installed, one command downloads and starts Qwen2.5 Coder 32B:

ollama run qwen2.5-coder:32b

Ollama's default download is Q4_K_M (about 22.6 GB here). You have room for FP16, which is full quality: pick that GGUF in LM Studio or from Hugging Face.

New to this? Step-by-step setup for the Apple M3 Ultra (256GB)

Run it in the cloud

No Apple M3 Ultra (256GB)? Run Qwen2.5 Coder 32B in the cloud.

Rent a cloud GPU by the hour. No upfront cost, stop any time.

Rent a GPU on RunPodRent a GPU on Vast.ai

Memory needed by quantization

At 8,192 tokens of context. The black line marks the 192 GB the Apple M3 Ultra (256GB) can use. macOS lets the GPU use about 192 GB of the 256 GB unified memory by default.

QuantQualityMemoryFits?Speed
FP16Full quality68.3 GBFits~7 tok/s
Q8_0Near-lossless37.6 GBFits~14 tok/s
Q6_KExcellent29.6 GBFits~18 tok/s
Q5_K_MVery good26.1 GBFits~21 tok/s
Q4_K_MGood, most popular22.6 GBFits~25 tok/s
Q3_K_MNoticeably degraded18.8 GBFits~31 tok/s
Q2_KHeavily degraded15 GBFits~40 tok/s

Measured speed from real Apple M3 Ultra (256GB) owners

Tokens per second that people measured on their own Apple M3 Ultra (256GB), next to our estimate.

No one has measured Qwen2.5 Coder 32B on the Apple M3 Ultra (256GB) yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an Apple M3 Ultra (256GB)? Add your real speed

One command times Qwen2.5 Coder 32B in your own Ollama, shows you exactly what it would send, and asks before sending anything. No sign-up, no personal data.

macOS and Linux

curl -fsSL https://gpufit.dev/bench.sh | sh -s -- qwen2.5-coder:32b

Windows (PowerShell)

$env:GPUFIT_MODEL="qwen2.5-coder:32b"; irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

GPUs that run Qwen2.5 Coder 32B comfortably

Models that run well on the Apple M3 Ultra (256GB)

Frequently asked questions

Can Apple M3 Ultra (256GB) run Qwen2.5 Coder 32B?

Yes. The Apple M3 Ultra (256GB) can run Qwen2.5 Coder 32B fully in unified memory at FP16 (full quality), using about 68.3 GB of the 192 GB available, at roughly 7 tokens/second.

How much VRAM does Qwen2.5 Coder 32B need?

About 22.6 GB at Q4_K_M with 8,192 tokens of context, and about 68.3 GB at full FP16 precision.

How fast is Qwen2.5 Coder 32B on the Apple M3 Ultra (256GB)?

Roughly 7 tokens per second at FP16, estimated from the Apple M3 Ultra (256GB)'s 819 GB/s memory bandwidth.