Qwen3.5-27B-TurboQuant-MLX-8bit

MLX 8-bit weight-quantized variant of Qwen/Qwen3.5-27B with TurboQuant KV cache compression for efficient inference on Apple Silicon.

Overview

This model combines two complementary compression techniques:

• MLX 8-bit weight quantization — reduces model size from ~54GB to ~27GB
• TurboQuant KV cache compression — compresses key-value caches during inference via online vector quantization, enabling longer contexts with less VRAM

Quickstart

from mlx_lm import load, generate
from turboquant import TurboQuantCache

model, tokenizer = load("majentik/Qwen3.5-27B-TurboQuant-MLX-8bit")

# Standard generation
prompt = "Explain the theory of relativity"
response = generate(model, tokenizer, prompt=prompt, max_tokens=2048)
print(response)

Specifications

Property	Value
Base Model	Qwen/Qwen3.5-27B
Parameters	27B
Weight Quantization	MLX 8-bit affine
KV Cache Method	TurboQuant (4-bit online vector quantization)
Model Size	~27 GB
Context Length	262K (native), 1M+ (extended)
Platform	Apple Silicon (M1/M2/M3/M4/M5)

What is TurboQuant?

TurboQuant (arXiv: 2504.19874) applies online vector quantization to key-value caches during inference, compressing them to 4-bit precision with near-lossless quality. This reduces the memory footprint of the KV cache by approximately 4x compared to FP16, allowing significantly longer context windows to fit in memory without retraining or fine-tuning.

Thinking Mode

Qwen3.5-27B generates extended reasoning before responses by default. The combination of weight quantization and KV cache compression is especially valuable here — thinking tokens consume significant memory that is reduced by both techniques working together.

Memory Estimate

Configuration	Model Weights	KV Cache (128K ctx)	Total
FP16 (baseline)	~54 GB	~13 GB	~67 GB
MLX 8-bit + TurboQuant	~27 GB	~3.3 GB	~30.3 GB

See Also

• Base model
• KV-cache-only variant
• TurboQuant paper (arXiv: 2504.19874)

Quant trade-off (MLX lane)

Bits	Approx size	Use case	Recommendation
2-bit	~7.3 GB	Aggressive quantization	Very low-RAM Macs
3-bit	~10 GB	Lossy but small	Low-RAM Macs
4-bit	~12 GB	Balanced default	Recommended for most Macs
5-bit	~14 GB	Higher fidelity	Quality-sensitive
6-bit	~17 GB	Approaching FP16 quality	High-fidelity
8-bit	~21 GB	Near-lossless reference	Fidelity-critical work

(Current variant — 8bit — is bolded.)

Variants in this family

(Showing 16 sibling variants under majentik/qwen3.5-27b-*. The current variant — TurboQuant-MLX-8bit — is bolded.)

Variant	Runtime	Approx size	Use case
RotorQuant	runtime modifier	n/a	KV-cache root (weight-agnostic)
RotorQuant-2bit	transformers	n/a	Standalone 2-bit weights
RotorQuant-GGUF-IQ4_XS	llama.cpp	~23 GB	Lossy 4-bit, low-RAM CPU/edge
RotorQuant-GGUF-Q2_K	llama.cpp	~16 GB	Lossy, low-RAM CPU/edge
RotorQuant-GGUF-Q3_K_M	llama.cpp	~21 GB	Smaller 3-bit, CPU-friendly
RotorQuant-GGUF-Q4_K_M	llama.cpp	~30 GB	Balanced default
RotorQuant-GGUF-Q5_K_M	llama.cpp	~36 GB	Higher fidelity, more RAM
RotorQuant-GGUF-Q8_0	llama.cpp	~57 GB	Near-lossless reference
RotorQuant-MLX-2bit	mlx-lm	~8.6 GB	Apple Silicon, smallest
RotorQuant-MLX-4bit	mlx-lm	~17 GB	Apple Silicon balanced
RotorQuant-MLX-8bit	mlx-lm	~32 GB	Apple Silicon reference
TurboQuant	runtime modifier	n/a	KV-cache root (weight-agnostic)
TurboQuant-2bit	transformers	n/a	Standalone 2-bit weights
TurboQuant-MLX-2bit	mlx-lm	~8.6 GB	Apple Silicon, smallest
TurboQuant-MLX-4bit	mlx-lm	~17 GB	Apple Silicon balanced
TurboQuant-MLX-8bit	mlx-lm	~32 GB	Apple Silicon reference