Run Phi-5 Large 28B on M4 Pro

Yes — Phi-5 Large 28B (28B) runs at 28 tok/s on M4 Pro with 32 GB RAM using Q4_K_M quantization via MLX. First token latency is 0.9s. A capable open-source LLM with 28B parameters.

Speed
28
tok/s
First Token
0.9
seconds
RAM Needed
32
GB minimum
Engine
MLX
recommended

Benchmark Details

LLMCheck measured Phi-5 Large 28B on M4 Pro using the standard methodology: Q4_K_M quantization, 256-token input, 512-token output, 3 runs averaged on a freshly-booted system.

MetricValue
Tokens per second28 tok/s
Time to first token0.9s
QuantizationQ4_K_M
Minimum RAM32 GB
Recommended engineMLX
Parameters28B
Benchmark date2026-07

Q4_K_M 28B MLX M4 Pro

Setup Guide: Run Phi-5 Large 28B on M4 Pro

The recommended engine for Phi-5 Large 28B on M4 Pro is MLX. Install with pip and pull the model:

pip install mlx-lm
mlx_lm.generate --model mlx-community/phi-5-large-28b-q4_k_m --prompt "Hello!"

Alternatively, you can use Ollama for a simpler setup:

ollama run phi-5-large-28b

Performance on Other Apple Silicon Chips

ChipSpeedFirst TokenMin RAMEngine
M5 Max 38 tok/s 0.6s 128 GB MLX
M5 Max 34 tok/s 0.7s 64 GB Ollama
M3 Max 26 tok/s 1.1s 64 GB Ollama

System Requirements

To run Phi-5 Large 28B on M4 Pro you need:

🛒 Get a Mac that runs Phi-5 Large 28B

Phi-5 Large 28B needs about 32 GB of unified memory. These current Apple Silicon Macs have the headroom to run it comfortably:

Not sure which Mac fits your budget? See the best Mac for running this →

As an Amazon Associate, LLMCheck earns from qualifying purchases. These affiliate links cost you nothing extra and help keep our benchmarks free.

Compare More Models

See how Phi-5 Large 28B stacks up against other models on your specific Mac hardware.

Open Compare Tool Full Leaderboard