Rent a GPU

Can the Apple M4 Max (64GB) run Phi-3.5 Mini 3.8B?

Apple M4 Max (64GB) × Phi-3.5 Mini 3.8B8K context
YesRuns fully on the GPU
Memory at FP1611.5 GB of 48 GB
Weights 7.6 GBKV cache 3.2 GBRuntime 0.6 GB
QuantFP16Full quality
Speed~43tokens per second
Max context98Ktokens that fit at FP16
  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 M4 Max (64GB) can run Phi-3.5 Mini 3.8B fully in unified memory at FP16 (full quality), using about 11.5 GB of the 48 GB available, at roughly 43 tokens/second.

Try other context lengths

Run it with Ollama

With Ollama installed, one command downloads and starts Phi-3.5 Mini 3.8B:

ollama run phi3.5:3.8b

Ollama's default download is Q4_K_M (about 6.1 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 M4 Max (64GB)

Run it in the cloud

No Apple M4 Max (64GB)? Run Phi-3.5 Mini 3.8B 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 48 GB the Apple M4 Max (64GB) can use. macOS lets the GPU use about 48 GB of the 64 GB unified memory by default.

QuantQualityMemoryFits?Speed
FP16Full quality11.5 GBFits~43 tok/s
Q8_0Near-lossless7.9 GBFits~81 tok/s
Q6_KExcellent7 GBFits~105 tok/s
Q5_K_MVery good6.5 GBFits~121 tok/s
Q4_K_MGood, most popular6.1 GBFits~141 tok/s
Q3_K_MNoticeably degraded5.7 GBFits~175 tok/s
Q2_KHeavily degraded5.3 GBFits~229 tok/s

Measured speed from real Apple M4 Max (64GB) owners

Tokens per second that people measured on their own Apple M4 Max (64GB), next to our estimate.

No one has measured Phi-3.5 Mini 3.8B on the Apple M4 Max (64GB) yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an Apple M4 Max (64GB)? Add your real speed

One command times Phi-3.5 Mini 3.8B 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 -- phi3.5:3.8b

Windows (PowerShell)

$env:GPUFIT_MODEL="phi3.5:3.8b"; irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

GPUs that run Phi-3.5 Mini 3.8B comfortably

Models that run well on the Apple M4 Max (64GB)

Frequently asked questions

Can Apple M4 Max (64GB) run Phi-3.5 Mini 3.8B?

Yes. The Apple M4 Max (64GB) can run Phi-3.5 Mini 3.8B fully in unified memory at FP16 (full quality), using about 11.5 GB of the 48 GB available, at roughly 43 tokens/second.

How much VRAM does Phi-3.5 Mini 3.8B need?

About 6.1 GB at Q4_K_M with 8,192 tokens of context, and about 11.5 GB at full FP16 precision.

How fast is Phi-3.5 Mini 3.8B on the Apple M4 Max (64GB)?

Roughly 43 tokens per second at FP16, estimated from the Apple M4 Max (64GB)'s 546 GB/s memory bandwidth.