Can I run Qwen 3.6 35B-A3B on a M6 16GB Mac?

Q: Can I run Qwen 3.6 35B-A3B on a M6 16GB Mac?

No — at 32K context Qwen 3.6 35B-A3B (8-bit) needs about 47.3 GB of the 16 GB unified memory on a M6 16GB Mac. Computed from the model's official config.json by the open-source FitLLM engine.

❌ Not at 32K context — Qwen 3.6 35B-A3B (8-bit) needs 47.3 GB of 16 GB unified memory

Computed with the open FitLLM engine — accurate per-layer KV-cache modeling, not a naive estimate. Updated 2026-10-19.

Conditions: 8-bit · KV cache F16 · 32K context · macOS reserve (9.0 GB).

Memory breakdown (8-bit, F16 KV, 32K context)

Model weights	32.6 GB
KV cache	0.6 GB
Linear-attention state (fixed — does not grow with context)	0.1 GB
Runtime + macOS	14.0 GB
Total used	47.3 / 16 GB
Short by	31.3 GB

Context at 8-bit: does not fit at any context. Unified memory is shared with macOS and your apps — the recommended max leaves 20% of it free.

Every quantization on M6 16GB

Quant	Weights	Recommended max ctx	Used @32K
4bit	16.3 GB	won't fit	❌ 29.0 / 16 GB
8bit	32.6 GB	won't fit	❌ 47.3 / 16 GB
16bit	65.2 GB	won't fit	❌ 83.8 / 16 GB

Recommended max = the longest context that still keeps the 20% headroom macOS needs · hard limit = the longest that fits at all (KV cache F16).

Lower quants free memory at some output-quality cost — 4-bit is the common sweet spot for local use.

▶ Open the interactive calculator (this exact setup)

Embed this verdict

Live badge for your README or model card — recomputed by the engine at exactly these conditions, never stale:

[![fits: Qwen 3.6 35B-A3B on M6 16GB Mac](https://img.shields.io/endpoint?url=https%3A%2F%2Ffitllm.run%2Fapi%2Fbadge%3Fmodel%3DQwen%2B3.6%2B35B-A3B%26ram%3D16%26quant%3D8%26ctx%3D32768%26kv%3D16)](https://fitllm.run/can-i-run/qwen-3-6-35b-a3b-on-m6-16gb)

fit badge preview ← renders like this, live.

Or from your terminal (exit 0/1 — works as a pre-download guard):

npx fitllm "Qwen 3.6 35B-A3B" --mac 16 --quant 8 --ctx 32768 --kv 16

Why most calculators get this wrong

Qwen 3.6 35B-A3B is a hybrid model: only 10 of its 40 layers use full attention — the rest are linear and keep no growing KV cache. Naive calculators count every layer at full context and badly over-estimate.

Other options

same Mac Models that fit in 16GB: MiniCPM5-1B, Qwen3-0.6B, Qwen3-1.7B, Llama-3.2-1B-Instruct, Gemma-3-1B-it, MiniCPM5-2B.

Reproduce it

Open math: fitllm-engine (MIT), from official config.json.

All numbers are computed by the open-source fitllm-engine (MIT) from official model config.json values — reproduce or audit them yourself. Estimates; real usage varies with runtime (llama.cpp / MLX / Ollama), driver and display. Found a mismatch? Report it. · FitLLM home