Run Qwen3.8 27B on Apple Silicon

Qwen3.8 27B is a dense 27 billion parameter model with image input and a 256K context, released by Alibaba's Qwen team in August 2026 under Apache 2.0. At Q4_K_M the weights are about 17 GB, so a Mac with 32 GB of unified memory runs it and a Mac mini M4 Pro with 64 GB runs it with long context to spare. Three of every four layers use linear attention, so the KV cache stays small as the context grows.

How much memory does Qwen3.8 27B need?

Weight size is parameter count × bits per weight; recommended memory adds 15% runtime headroom plus a KV-cache allowance for the 262,144-token context window.

Which Mac should you pick?

Generation speed on Apple Silicon scales mostly with memory bandwidth, so the chip choice sets both whether the model fits and how fast it runs.

Deploy Qwen3.8 27B in about 5 minutes

1. 1

   Build a Mac in the constructor — pick an M4 with at least 24 GB of unified memory.

2. 2

   Pick any Ollama-based stack (or a clean machine) and pull Qwen3.8 27B with one command — every deployment ships the OpenAI-compatible API.

3. 3

   Point your existing OpenAI SDK code at the new base_url and ship.

from openai import OpenAI

client = OpenAI(
    api_key="mcy_live_...",
    base_url="https://dep-xxxx.macyou.cloud/v1",
)
response = client.chat.completions.create(
    model="qwen-3-8-27b",
    messages=[{"role": "user", "content": "Hello!"}],
)

License and what it's good at

Apache 2.0 — Apache 2.0: commercial use is allowed under the standard license terms. Made by Alibaba (Qwen).

• Apache 2.0 license
• Text and image input with a 256K context
• Fits a 32 GB Mac at Q4

Typical workloads: a private daily assistant and coding helper on one mac; reading documents, screenshots and charts; a default model for agents that need a long context.

Frequently asked questions

Related

• Run Qwen3.8 Flash Next on Apple Silicon →
• Run Gemma 4 31B on Apple Silicon →
• Buying vs renting a Mac for AI →