Run Phi-5 Medium 14B on M3

Yes — Phi-5 Medium 14B (14B) runs at 32 tok/s on M3 with 16 GB RAM using Q4_K_M quantization via Ollama. First token latency is 1.1s. A capable open-source LLM with 14B parameters.

Speed
32
tok/s
First Token
1.1
seconds
RAM Needed
16
GB minimum
Engine
Ollama
recommended

Benchmark Details

LLMCheck measured Phi-5 Medium 14B on M3 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 second32 tok/s
Time to first token1.1s
QuantizationQ4_K_M
Minimum RAM16 GB
Recommended engineOllama
Parameters14B
Benchmark date2026-06

Q4_K_M 14B Ollama M3

Setup Guide: Run Phi-5 Medium 14B on M3

The recommended engine for Phi-5 Medium 14B on M3 is Ollama. Install Ollama, then pull the model:

ollama run phi-5-medium-14b

Ollama handles quantization automatically — it will download the Q4_K_M variant (~16 GB) and start an interactive chat session.

Performance on Other Apple Silicon Chips

ChipSpeedFirst TokenMin RAMEngine
M5 Max 65 tok/s 0.5s 128 GB MLX
M5 Max 58 tok/s 0.6s 64 GB Ollama
M4 Pro 48 tok/s 0.7s 24 GB MLX
M2 24 tok/s 1.4s 16 GB Ollama

System Requirements

To run Phi-5 Medium 14B on M3 you need:

Compare More Models

See how Phi-5 Medium 14B stacks up against other models on your specific Mac hardware.

Open Compare Tool Full Leaderboard