What are the hardware requirements for Swift 1.5 Qwen3.8 27B AWQ BF16 INT4?

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 requires approximately 28.8 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 and how large is it?

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 is developed by cyankiwi, a model with 27b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Swift 1.5 Qwen3.8 27B AWQ BF16 INT4?

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 by cyankiwi — VRAM 28.8GB

Name: Swift 1.5 Qwen3.8 27B AWQ BF16 INT4
Author: cyankiwi

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 is an open-source language model by cyankiwi. Features: 27b LLM, VRAM: 28.8GB, License: other, Quantized, LLM Explorer Score: 0.32.

Awq Base model:quantized:ukisai/sw... Base model:ukisai/swift-1.5-qw... Compressed-tensors Conversational Efficient-thinking Endpoints compatible Image-text-to-text Post-training Quantized Qwen3 5 Qwen3 8 Reasoning Region:us Safetensors Sharded Tensorflow Terminal-bench Token-efficient

Model Card on HF 🤗: https://huggingface.co/cyankiwi/Swift-1.5-Qwen3.8-27B-AWQ-BF16-INT4

Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 Parameters and Internals

LLM Name	Swift 1.5 Qwen3.8 27B AWQ BF16 INT4
Repository 🤗	https://huggingface.co/cyankiwi/Swift-1.5-Qwen3.8-27B-AWQ-BF16-INT4
Base Model(s)	ukisai/Swift-1.5-Qwen3.8-27b ukisai/Swift-1.5-Qwen3.8-27b
Model Size	27b
Required VRAM	28.8 GB
Updated	2026-10-05
Maintainer	cyankiwi
Model Type	qwen3_5
Model Files	2.5 GB: 1-of-7 5.0 GB: 2-of-7 5.0 GB: 3-of-7 4.9 GB: 4-of-7 4.9 GB: 5-of-7 5.0 GB: 6-of-7 1.5 GB: 7-of-7
AWQ Quantization	Yes
Quantization Type	awq
Model Architecture	Qwen3_5ForConditionalGeneration
License	other
Model Max Length	262144
Transformers Version	5.8.0.dev0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the Swift 1.5 Qwen3.8 27B AWQ BF16 INT4 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Swift 1.5 Qwen3.8 27B AWQ BF16 INT4

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3.6 27B AWQ INT4	0K / 20.4 GB	2563390	103
Qwen3.8 27B AWQ INT4	0K / 21 GB	21015	32
Qwen3.6 27B AWQ	0K / 22 GB	668924	24
Qwen3.6 27B AWQ BF16 INT4	0K / 28.3 GB	419171	41
...wen3.6 27B Abliterated AWQ MTP	0K / 18.4 GB	244978	12
Qwen3.5 27B AWQ	0K / 21.9 GB	370920	44
Qwen3.5 27B AWQ 4bit	0K / 20 GB	279064	42
Swift 1.5 Qwen3.8 27B AWQ INT4	0K / 21 GB	550	5
Qwen3.8 27B AWQ W4A16 ASYM	0K / 0.8 GB	13912	3
Qwen3.8 27B AWQ 4bit	0K / 27.7 GB	11874	6

Note: green Score (e.g. "73.2") means that the model is better than cyankiwi/Swift-1.5-Qwen3.8-27B-AWQ-BF16-INT4.