What are the hardware requirements for SpatialBlock 4B Direct?

SpatialBlock 4B Direct requires approximately 8.9 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SpatialBlock 4B Direct and how large is it?

SpatialBlock 4B Direct is developed by rsoohyun, 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 SpatialBlock 4B Direct?

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

SpatialBlock 4B Direct by rsoohyun — VRAM 8.9GB

Name: SpatialBlock 4B Direct
Author: rsoohyun

SpatialBlock 4B Direct is an open-source language model by rsoohyun. Features: 4b LLM, VRAM: 8.9GB, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.25.

Arxiv:2609.07064 Base model:finetune:qwen/qwen3... Base model:qwen/qwen3-vl-4b-in... Conversational Dataset:rsoohyun/spatialblock-... Endpoints compatible Image-text-to-text Instruct Qwen3 vl Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/rsoohyun/SpatialBlock-4B-direct

SpatialBlock 4B Direct Parameters and Internals

LLM Name	SpatialBlock 4B Direct
Repository 🤗	https://huggingface.co/rsoohyun/SpatialBlock-4B-direct
Base Model(s)	Qwen3 VL 4B Instruct Qwen/Qwen3-VL-4B-Instruct
Model Size	4b
Required VRAM	8.9 GB
Updated	2026-09-13
Maintainer	rsoohyun
Model Type	qwen3_vl
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 3.9 GB: 2-of-2
Model Architecture	Qwen3VLForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	4.57.6
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the SpatialBlock 4B Direct 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 SpatialBlock 4B Direct

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3 VL 4B Instruct	0K / 8.9 GB	3611570	437
MinerU Popo	0K / 17.6 GB	26895	4
Qwen3 VL 4B Instruct FP8	0K / 6.1 GB	301103	65
AA VL 4B	0K / 8.9 GB	87	0
Qwen3 VL 4B VQA SFT	0K / 8.9 GB	40	0
Kayak V2	0K / 17.7 GB	31	0
Op Hrg Qwen3vl4b Instruct	0K / 9.7 GB	51	2
Docision 1.0	0K / 10.5 GB	30	1
Slayer Ocr Qwen3vl 4B Frozen	0K / 8.9 GB	39	0
MJev Qwen3 VL 4B RLCD	0K / 9.7 GB	40	2

Note: green Score (e.g. "73.2") means that the model is better than rsoohyun/SpatialBlock-4B-direct.