How to Run Qwen3.6 Locally — Setup Guide (August 2026)

Qwen3.6 is Alibaba's current flagship open-weight family — a dense 27B and a 35B-A3B Mixture-of-Experts sibling that activates only ~3B parameters per token. Both ship as GGUF, run on llama.cpp (or the TurboQuant fork, for a smaller KV cache), and are the exact models behind this site's own published benchmarks.

Numbers you can trust

We don't print guessed speeds. Every measured number on this page came from TurboLLM's own auto-tuner running Qwen3.6 on real hardware — labeled as such. When you run it yourself, TurboLLM benchmarks it on your exact GPU and shows a VRAM-fit verdict before you load, and the real measured tokens/sec once it's running.

Does it fit your GPU?

The 35B-A3B variant is a Mixture-of-Experts model with a hybrid attention design — only a fraction of its layers keep a growing KV cache, while the rest use a constant-size recurrent state. That's why it can hold a 200K-token context without the KV cache ballooning the way a plain dense model's would.

Qwen3.6 27B
Alibaba · GGUF
Dense
toolsthinkingcode
VRAMFit
16 GBIQ4_XS (15.4 GB) or Q3_K_M (13.6 GB) in VRAM · Q4_K_M (16.8 GB) + light offload · 256K context
24 GB+Fully in VRAM at Q4_K_MQ5_K_M with room for long context · 256K context
unsloth/Qwen3.6-27B-GGUF
Qwen3.6 35B-A3B
Alibaba · GGUF
MoE · ~3B active
toolsthinkingcode
VRAMFit
16 GBQ3_K_M (16.6 GB) just spills past 16 GB — auto-fit nudges a few experts to system RAM — or Q4_K_M (22 GB) + expert offload · 256K context
24 GB+Q4_K_M (22 GB) fully in VRAM with room for context, or Q5_K_M (26.5 GB) + light offload · 256K context
48 GBQ8_0 (37 GB) fully in VRAM · 256K context
unsloth/Qwen3.6-35B-A3B-GGUF

How to run it

  1. Install TurboLLM

    npx turbollm — no install step, works on Windows, macOS, and Linux. It detects your GPU and auto-provisions a matching llama-server build (no CUDA toolkit, no Python, no compiler).

  2. Get the model

    Paste unsloth/Qwen3.6-27B-GGUF into TurboLLM's in-app Hugging Face search — it lists every quant with a VRAM-fit verdict against your real free VRAM, and downloads are resumable and SHA-256 verified.

  3. Let auto-fit pick the setup

    Flip the auto-fit toggle and TurboLLM chooses the GPU/CPU layer split (and, for a Mixture-of-Experts model, the expert-offload split) for your card — the same manual flag-hunting every other guide walks you through by hand.

  4. Load it

    TurboLLM benchmarks the model on your exact GPU as it loads and shows the real measured tokens/sec — not a number copied from someone else's card.

  5. Use it

    Chat in the built-in UI, or point any OpenAI- or Anthropic-compatible tool (including Claude Code) at TurboLLM's local API — see the API docs.

Need the TurboQuant fork instead?

TurboQuant's turbo4 KV-cache format doesn't exist in LM Studio's bundled runtime. Being able to load a community fork through the same engines catalog — no compiling — is what turns the 27B's KV-cache bottleneck into the 2.2× row above.

Measured, not guessed

RTX 5070 Ti 16 GBTurboLLMLM Studio
Qwen3.6-35B-A3B · official llama.cpp, q4_0 KV, 200K ctx74.7 t/s61.0 t/s
Qwen3.6-35B-A3B · official llama.cpp, q8_0 KV, 200K ctx72.3 t/s~66 t/s
Qwen3.6-27B · TurboQuant turbo4 KV (LM Studio: q8_0 KV), 200K ctx24.6 t/s (2.2×)11.4 t/s
Qwen3.6-27B · prefill (prompt processing)1288 t/s757 t/s

Founder-measured on an RTX 5070 Ti 16 GB, July 2026 — same machine, same GGUF files. Full methodology on the TurboLLM vs LM Studio page. These numbers also appear on the GPU page for this card.

TurboQuant's turbo4 KV-cache format doesn't exist in LM Studio's bundled runtime. Being able to load a community fork through the same engines catalog — no compiling — is what turns the 27B's KV-cache bottleneck into the 2.2× row above.

FAQ

Why does Qwen3.6 suddenly get slow or stutter?

The usual cause is silent VRAM spill: once a model's weights plus its KV cache stop fitting in VRAM, the driver quietly offloads the overflow to system RAM over PCIe, and that path runs 5–10× slower with no error message — just a sudden slowdown. TurboLLM's auto-fit sizes the quant and GPU/CPU split against your card's real free VRAM (including KV-cache growth at your chosen context) specifically to avoid this, and shows a VRAM-fit verdict before you load rather than after it's already crawling. Full mechanics in VRAM spill explained.

Which quant should I use for Qwen3.6?

Pick the highest quant that fits your VRAM with room left over for the KV cache and compute buffer — not just the model weights. As a starting point, Q4_K_M is usually the sweet spot (small quality loss for a large size drop); step up to Q5_K_M/Q6_K/Q8_0 if you have headroom, or down to Q3_K_M/an IQ quant if you don't. TurboLLM shows a VRAM-fit verdict for every quant against your card's real free VRAM before you download — see Quantization explained for what the letters and numbers actually mean, or llama.cpp flags that actually matter for the other settings worth knowing.

Will Qwen3.6 run on my GPU?

Check the VRAM table above for the tier closest to your card. If you're not sure which tier you're in, or want picks tailored to your exact hardware, use What can I run? — enter your VRAM (or Apple unified memory) and it suggests the model, quant, and offload split that fits, the same logic TurboLLM's auto-fit runs when you load a model.

How is running Qwen3.6 on TurboLLM different from Ollama or LM Studio?

All three can run Qwen3.6 as a GGUF. The differences show up around the edges: TurboLLM auto-tunes the quant and GPU/CPU split against your card's measured free VRAM rather than a fixed default, shows real generation speed instead of a guess, and lets you add a community fork — like the one behind this page, where relevant — without compiling anything yourself. Ollama and LM Studio each ship one blessed runtime you can't swap out. If your current setup already works well for you, there's no need to switch.

Run it now

$ npx turbollm

One command detects your GPU, provisions the right engine, and opens the UI. New here? Start with Install & first run and Quantization explained. On different hardware, see RTX 5070 Ti, RTX 5080, RTX 5060 Ti or RTX 4090. Also see how to run it.