Qwen3.8-27B

· llama.cpp · 128K context · 2 × RTX 3060 in one machine · 24 GB

Alibaba's 27B dense model from August 2026, and the strongest model that fits one consumer card. It thinks before it answers, calls tools reliably and reads images, which makes it the default for coding agents and long projects.

56 tok/sdecode

497 tok/sprefill

128Kcontext window

Tested on this cardOct 1, 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 unsloth/Qwen3.8-27B-GGUF Qwen3.8-27B-UD-Q4_K_XL.gguf \
  --revision 4ca720788d1e01f1bff70c033e0d0028fd02e502 \
  --local-dir ~/models/Qwen3.8-27B-GGUF-4ca72078

2 Start the server

docker run --rm \
  --gpus all \
  -p 8000:8080 \
  --shm-size 16g \
  -e CUDA_DEVICE_ORDER=PCI_BUS_ID \
  -e GGML_CUDA_ALLREDUCE=internal \
  -v ~/models/Qwen3.8-27B-GGUF-4ca72078:/models:ro \
  --entrypoint /app/llama-server \
  ghcr.io/ggml-org/llama.cpp:server-cuda12-b11146@sha256:52192a23c7b258044463aebba78826cdff39a7393b713edc313324b343a7b836 \
  --model /models/Qwen3.8-27B-UD-Q4_K_XL.gguf \
  --n-gpu-layers 999 \
  --split-mode tensor \
  --tensor-split 1,1 \
  --cache-type-k q4_0 \
  --cache-type-v q4_0 \
  --host 0.0.0.0 \
  --port 8080 \
  --ctx-size 131072 \
  --flash-attn on \
  --parallel 1 \
  --jinja \
  --no-mmproj \
  --metrics \
  --spec-type draft-mtp \
  --spec-draft-n-max 3 \
  --temp 1.0 \
  --top-p 0.95 \
  --top-k 20 \
  --min-p 0.0

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 3060 on Oct 1, 2026.

Details

Weights: unsloth/Qwen3.8-27B-GGUF @ 4ca720788d ›
Image: ghcr.io/ggml-org/llama.cpp:server-cuda12-b11146@sha256:52192a23c7b258044463aebba78826cdff39a7393b713edc313324b343a7b836
Launch: llama.cpp-qwen3.8-27b-ud-q4-k-xl-tensor-mtp-128k-2x ›
Recipe file: registry/recipes/nvidia/rtx-3060-12gb/qwen3.8-27b.llama.cpp.128k.2x.json ›
Model released: Aug 5, 2026
Tested: owner, RTX 3060