LLM EXPLORER 62,589 MODELS INDEXED

Swift 1.5 4bit MLX by ukisai

By ukisai · 1333 downloads
HF Repo ↗ Copy URL

Swift 1.5 4bit MLX is an open-source language model by ukisai. Features: 27b LLM, VRAM: 15.7GB, License: other, Quantized, LLM Explorer Score: 0.34.

  4-bit   4bit   Affine Base model:quantized:ukisai/sw... Base model:ukisai/swift-1.5-qw...   Conversational   Mlx   Quantized   Qwen3 5   Qwen3 8   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/ukisai/Swift-1.5-4bit-MLX 

Swift 1.5 4bit MLX Parameters and Internals

LLM NameSwift 1.5 4bit MLX
Repository 🤗https://huggingface.co/ukisai/Swift-1.5-4bit-MLX 
Base Model(s)  ukisai/Swift-1.5-Qwen3.8-27b   ukisai/Swift-1.5-Qwen3.8-27b
Model Size27b
Required VRAM15.7 GB
Updated2026-09-30
Maintainerukisai
Model Typeqwen3_5
Model Files  5.3 GB: 1-of-3   5.3 GB: 2-of-3   5.1 GB: 3-of-3
Quantization Type4bit
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseother
Model Max Length262144
Transformers Version5.8.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Swift 1.5 4bit MLX

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.6 27B OptiQ 4bit256K / 18.7 GB1306559
Bonsai 27B Mlx 1bit0K / 5.1 GB809309205
Ternary Bonsai 27B Mlx 2bit0K / 8.5 GB809024168
Qwen3.8 27B MLX 5bit0K / 19.5 GB46750442
Qwen3.6 27B MLX 8bit0K / 29.4 GB87914514
Qwen3.6 27B MLX 6bit0K / 22.7 GB8129772
...onsai 2 27B CRACK Ternary JANG0K / 8.1 GB268913
Qwen3.6 27B MLX 4bit0K / 16 GB84686116
...ngCap Qwen3.8 27B MLX 4bit DWQ0K / 21.1 GB6749
Qwen3.8 27B 8bit0K / 29.4 GB1389233
Note: green Score (e.g. "73.2") means that the model is better than ukisai/Swift-1.5-4bit-MLX.