What are the hardware requirements for SmolVLM 256M Instruct?

SmolVLM 256M Instruct requires approximately 0.5 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SmolVLM 256M Instruct and how large is it?

SmolVLM 256M Instruct is developed by HuggingFaceTB, a model with 135m 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 SmolVLM 256M Instruct?

SmolVLM 256M Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

SmolVLM 256M Instruct by HuggingFaceTB — VRAM 0.5GB

Name: SmolVLM 256M Instruct
Author: HuggingFaceTB

SmolVLM 256M Instruct is an open-source language model by HuggingFaceTB. Features: 135m LLM, VRAM: 0.5GB, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.28.

Arxiv:2504.05299 Base model:huggingfacetb/smoll... Base model:quantized:huggingfa... Conversational Dataset:huggingfacem4/docmatix Dataset:huggingfacem4/the caul... En Endpoints compatible Idefics3 Image-text-to-text Instruct Onnx Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/HuggingFaceTB/SmolVLM-256M-Instruct

SmolVLM 256M Instruct Parameters and Internals

LLM Name	SmolVLM 256M Instruct
Repository 🤗	https://huggingface.co/HuggingFaceTB/SmolVLM-256M-Instruct
Base Model(s)	SmolLM2 135M Instruct google/siglip-base-patch16-512 HuggingFaceTB/SmolLM2-135M-Instruct google/siglip-base-patch16-512
Model Size	135m
Required VRAM	0.5 GB
Updated	2026-08-09
Maintainer	HuggingFaceTB
Model Type	idefics3
Instruction-Based	Yes
Model Files	0.5 GB
Supported Languages	en
Model Architecture	Idefics3ForConditionalGeneration
License	apache-2.0
Model Max Length	8192
Transformers Version	4.46.0
Tokenizer Class	GPT2Tokenizer
Padding Token	<\|im_end\|>
Vocabulary Size	49280
Torch Data Type	bfloat16

Rank the SmolVLM 256M Instruct 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 SmolVLM 256M Instruct

Best Alternatives	Context / RAM	Downloads	Likes
SmolVLM 256M Instruct Bf16	0K / 0.5 GB	79	1