What are the hardware requirements for ASL 4B V1?

ASL 4B V1 requires approximately 9.3 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed ASL 4B V1 and how large is it?

ASL 4B V1 is developed by saai-sa, a model with 4b 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 ASL 4B V1?

ASL 4B V1 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

ASL 4B V1 by saai-sa — VRAM 9.3GB

Name: ASL 4B V1
Author: saai-sa

ASL 4B V1 is an open-source language model by saai-sa. Features: 4b LLM, VRAM: 9.3GB, License: apache-2.0, LLM Explorer Score: 0.35.

Ar Conversational Qwen3 5 Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/saai-sa/ASL-4B-v1

ASL 4B V1 Parameters and Internals

LLM Name	ASL 4B V1
Repository 🤗	https://huggingface.co/saai-sa/ASL-4B-v1
Base Model(s)	Qwen3.5 /Qwen3.5
Model Size	4b
Required VRAM	9.3 GB
Updated	2026-08-28
Maintainer	saai-sa
Model Type	qwen3_5
Model Files	5.3 GB: 1-of-2 4.0 GB: 2-of-2
Supported Languages	ar
Model Architecture	Qwen3_5ForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	4.57.0.dev0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|vision_pad\|>
Errors	replace

Rank the ASL 4B V1 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 ASL 4B V1

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3.5 4B	0K / 9.3 GB	6575502	801
NuExtract3	0K / 0.2 GB	441570	324
Agents A1 4B	0K / 9.1 GB	177578	72
Qwen3.8 4B Distill	0K / 9.3 GB	1878	26
Qwen3.5 4B Base	0K / 9.3 GB	227814	79
Fara1.5 4B	0K / 9.1 GB	4934	39
AREX Turbo	0K / 9.1 GB	1071	45
Qwen3.5 4B	0K / 9.3 GB	139961	26
DeepSeek V4 Pro Qwen3.5 4B	0K / 9.3 GB	3476	9
Mephisto 4B 0725	0K / 8.6 GB	3121	6

Note: green Score (e.g. "73.2") means that the model is better than saai-sa/ASL-4B-v1.