Qwen3.6-35B-A3B

NVFP4 · vllm · 64K context · RTX PRO 4500 Blackwell · 32 GB

Qwen3.6's 35B mixture-of-experts model with 3B active parameters.

188 tok/sdecode

7,733 tok/sprefill

64Kcontext window

Tested on this cardSep 27, 2026

Run it

Follow the steps below to download and run this recipe. Omarchy Local AI lists the recipes it can launch on your hardware. The server then answers on http://localhost:8000/v1.

1 Download the weights

hf download nvidia/Qwen3.6-35B-A3B-NVFP4 \
  --revision 1355db6a052410cfd62085d94b58866fd0f2c3c5 \
  --local-dir ~/models/Qwen3.6-35B-A3B-NVFP4-1355db6a

2 Start the server

docker run --rm \
  --gpus all \
  -p 8000:8000 \
  --shm-size 16g \
  -v ~/models/Qwen3.6-35B-A3B-NVFP4-1355db6a:/models/Qwen3.6-35B-A3B-NVFP4-1355db6a:ro \
  --entrypoint vllm \
  vllm/vllm-openai:v0.30.0@sha256:8a69ffad015f138d7170c4ddc429e230a3bc1c1719f67e14324749df200a4b90 \
  serve \
  /models/Qwen3.6-35B-A3B-NVFP4-1355db6a \
  --served-model-name Qwen3.6-35B-A3B-NVFP4-1355db6a \
  --host 0.0.0.0 \
  --port 8000 \
  --max-model-len 65536 \
  --max-num-seqs 1 \
  --gpu-memory-utilization 0.92 \
  --kv-cache-dtype auto \
  --enable-auto-tool-choice \
  --tool-call-parser qwen3_coder \
  --reasoning-parser qwen3

What it passed

✓ loadLoads and serves within an hour

✓ chatAnswers a plain question and stops on its own

✓ reasoningThinks separately and gets 17 × 23 right

✓ toolsCalls a tool with the right arguments and uses the result

✓ contextRecalls a code buried in a prompt that fills 85% of the window

✓ speedDecodes at 15 tok/s or more

Passed all six checks on a real RTX PRO 4500 on Sep 27, 2026.

Details

Weights: nvidia/Qwen3.6-35B-A3B-NVFP4 @ 1355db6a05 ›
Image: vllm/vllm-openai:v0.30.0@sha256:8a69ffad015f138d7170c4ddc429e230a3bc1c1719f67e14324749df200a4b90
Engine profile: vllm ›
Recipe file: registry/recipes/nvidia/rtx-pro-4500-blackwell-32gb/qwen3.6-35b-a3b.vllm.64k.json ›
Model released: Apr 15, 2026
Tested: vast, RTX PRO 4500