blockblockblock/Qwen3.6-35B-A3B-exl3-5.0bpw

Qwen3.6 · 35B-A3B

_{EXL3  ·  5.0 bpw  ·  23.7 GB  ·  Mixture‑of‑Experts  ·  40 layers × 256 experts}

---

An ExLlamaV3 build of Qwen/Qwen3.6-35B-A3B at 5.0 bits per weight — the quality build — fits a 24 GB card with reduced context, ideal headroom on 32 GB cards. See Quants for sibling repos at other bit‑widths or browse the collection.

Quants

BPW	Head bits	Calibration rows	Size	Status
5.0	8	250	23.7 GB	this repo

Inference

Loader	Use it for
TabbyAPI	OpenAI‑compatible HTTP server. Drop‑in for OpenAI clients.
text‑generation‑webui	Local chat UI. Pick the ExLlamaV3 loader from the model dropdown.
ExLlamaV3	Direct Python API for embedding the model in your own code or pipeline.

VRAM at 5.0 bpw: weights on disk + ~2 GB context overhead. Tight on 24 GB (limited context); comfortable on 32 GB+.

Download

pip install -U huggingface_hub

hf download \
  blockblockblock/Qwen3.6-35B-A3B-exl3-5.0bpw \
  --local-dir ./Qwen3.6-35B-A3B-exl3-5.0bpw

Quantization recipe  _{(advanced — embedded in quantization_config.json)}

Setting	Value
Format	EXL3
Bits per weight	5.0
Head bits	8
Calibration rows	250
Codebook	MCG
Out‑scales	always
Parallel mode	enabled (MoE expert batching)

Loaded automatically by every ExLlamaV3 loader; reproduced here for searchability.

License & use

Use and license follow the base model. Quantization adds no additional restrictions. Refer to the upstream repository for terms, citation, and safety documentation.

---

_{Quantized with BlockQuant  ·  convention {org}/{model}-exl3-{bpw}bpw}