Qwen3.5 35B A3B (Q2_K_XL)on CPU Only

Q2_K_XL CPU Only

Overview

Qwen3.5 35B A3B is a 35.95B parameter moe language model by Qwen, with code, multilingual, thinking, tool-calls, vision capabilities. It supports a context window of up to 262,144 tokens.

Qwen3.5 35B A3B is a Mixture-of-Experts model from Alibaba's Qwen team with 35 billion total parameters but only 3 billion active per token, routed across 256 experts for extreme efficiency. It is natively multimodal, processing text, images, and video, with built-in thinking capabilities for chain-of-thought reasoning. The model supports a 262K context window and covers over 200 languages. Released under the Apache 2.0 license, it delivers flagship-level performance at a fraction of the compute cost, quantizing efficiently for self-hosted deployment on consumer hardware.

At Q2_K_XL quantization (low quality tier), the model weighs 12.04 GB. This exceeds the 0 GB of VRAM on CPU Only. Inference is still possible via CPU offload or memory-mapped loading from disk, but expect significantly reduced performance.

A CPU-only configuration with no GPU acceleration. Inference runs entirely on the CPU, which is significantly slower than GPU-accelerated setups but requires no special hardware. Performance and maximum model size depend on available system RAM. Suitable for testing, development, or deployments where no GPU is available.

Hardware Requirements

Model size	12.04 GB
VRAM available	0 GB
VRAM used	0 GB
Min RAM required	12 GB
GPU layers	0 / 40
Context size	262,144
Backend	cpu
Flash attention	No

Performance Notes

Mixture-of-Experts

This is a Mixture-of-Experts model with 256 total experts, of which 8 are active per token. Only the active experts need fast GPU access — inactive expert weights can reside in system RAM with minimal impact on generation speed.

Read the MoE deployment guide →

Reduced performance

40 of 40 layers run on CPU. Your system needs at least 12 GB RAM. If system RAM is insufficient, data will be read from disk on demand, with significant performance impact.

Deploy

Server Local

Command

helmfile --state-values-file <(curl -s https://www.prositronic.eu/values/qwen3-5-35b-a3b/q2_k_xl/cpu.yaml) apply

Generated values.yaml

/values/qwen3-5-35b-a3b/q2_k_xl/cpu.yaml

Loading values…

Install llama.cpp

Install llama.cpp from the official build instructions for your platform and backend.

Download Model

curl -L -o qwen3-5-35b-a3b.gguf "https://huggingface.co/unsloth/Qwen3.5-35B-A3B-GGUF/resolve/main/Qwen3.5-35B-A3B-UD-Q2_K_XL.gguf"

Start Server

llama-server \
  -m qwen3-5-35b-a3b.gguf \
  --n-gpu-layers 0 \
  --ctx-size 262144

Verify

curl http://localhost:8080/health

Explore

• Configure a deployment
• Browse models
• Browse hardware

Other quantizations

• Qwen3.5 35B A3B (Q4_K_M)

Other hardware

• Qwen3.5 35B A3B on NVIDIA A100 80GB
• Qwen3.5 35B A3B on NVIDIA H100
• Qwen3.5 35B A3B on NVIDIA RTX 3090

Frequently Asked Questions

How much VRAM does Qwen3.5 35B A3B (Q2_K_XL) need?

The Q2_K_XL quantization of Qwen3.5 35B A3B requires 12.04 GB. The 0 GB of VRAM on CPU Only is insufficient for GPU layers, so inference runs on CPU.

Can I run Qwen3.5 35B A3B on CPU Only?

It is possible but not recommended. CPU Only does not have enough VRAM to accelerate Qwen3.5 35B A3B (Q2_K_XL), so inference will rely on CPU and system RAM.

What is quantization?

Quantization reduces a model's numerical precision from its original floating-point format to a more compact representation. This shrinks the file size and VRAM footprint, making it possible to run large models on consumer hardware. The trade-off is a small reduction in output quality. Q2_K_XL compresses Qwen3.5 35B A3B from its original size down to 12.04 GB.

What quantization should I choose for Qwen3.5 35B A3B?

Q2_K_XL is a low-quality quantization. Higher-quality quants (Q8, Q6) preserve more model accuracy but need more VRAM. Lower quants (Q4, Q3, Q2) reduce VRAM usage at the cost of some quality. Choose based on your available hardware and quality requirements.

Why are some layers offloaded to CPU?

CPU Only has 0 GB of VRAM, but Qwen3.5 35B A3B (Q2_K_XL) requires approximately 12.04 GB. Only 0 of 40 layers fit in VRAM; the remaining layers run on CPU, which is slower but still functional.

What is MoE and how does it affect deployment?

Qwen3.5 35B A3B uses a Mixture-of-Experts (MoE) architecture with 256 experts, of which 8 are active per token. This means only a fraction of the model weights are used for each inference step, allowing MoE models to be larger in total parameter count while remaining efficient at inference time.

How do I run Qwen3.5 35B A3B (Q2_K_XL) with Ollama?

Run ollama run qwen3.5:35b-a3b-q2_k_xl to start Qwen3.5 35B A3B (Q2_K_XL). Ollama handles downloading the model weights automatically on first run.

Last updated: March 13, 2026