Smaller in size (~0.9 GB) for the same quant level as the multimodal version, better suited for agentic use and can sustain multiple concurrent streams.

Supports {%- set enable_thinking = false %} Jinja variable.

Update [2026-04-12]: Refined the chat template to further improve stability for long-running tasks and tool use; mitigated an issue where incorrect think tag formatting could leak from a distillation dataset.

Quantization Details

Property	Value
Method	5-bit (5.501 bits per weight)
Tool	`mlx-lm 0.31.1` via `mlx-lm.convert`
Size	~18.5GB

Other Available Quants

Model	Size	Quantization	Bits per weight	Multimodal
Jackrong/MLX-Qwopus3.5-27B-v3-4bit	15.15 GB	4-bit	4.501	✗
matt-here/MLX-Qwopus3.5-27B-v3-vision-4bit	16.08 GB	4-bit	4.695	✓ (Vision)
(This model)	18.56 GB	5-bit	5.501	✗
matt-here/MLX-Qwopus3.5-27B-v3-vision-5bit	19.46 GB	5-bit	5.678	✓ (Vision)
Jackrong/MLX-Qwopus3.5-27B-v3-6bit	21.88 GB	6-bit	6.501	✗
matt-here/MLX-Qwopus3.5-27B-v3-vision-6bit	22.85 GB	6-bit	6.661	✓ (Vision)
Jackrong/MLX-Qwopus3.5-27B-v3-bf16	53.81 GB	bf16	16	✗

GGUF quants - Jackrong/Qwopus3.5-27B-v3-GGUF

Credits

Alibaba Qwen Team — Qwen 3.5 27B dense model
Jackrong - Claude 4.6 Opus v3 distillation work
Unsloth - Training framework
Apple MLX Team - High-speed local inference on Apple Silicon

matt-here/MLX-Qwopus3.5-27B-v3-5bit

MLX-Qwopus3.5-27B-v3-5bit

Quantization Details

Other Available Quants

Credits

matt-here/
MLX-Qwopus3.5-27B-v3-5bit