What are the hardware requirements for Laguna XS 2.1 Vision NVFP4?

Laguna XS 2.1 Vision NVFP4 requires approximately 21.5 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Laguna XS 2.1 Vision NVFP4 and how large is it?

Laguna XS 2.1 Vision NVFP4 is developed by webbrain-one, a model with 33.4b 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 Laguna XS 2.1 Vision NVFP4?

Laguna XS 2.1 Vision NVFP4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Laguna XS 2.1 Vision NVFP4 by webbrain-one — VRAM 21.5GB, 256K context

Name: Laguna XS 2.1 Vision NVFP4
Author: webbrain-one

Laguna XS 2.1 Vision NVFP4 is an open-source language model by webbrain-one. Features: 33.4b LLM, VRAM: 21.5GB, Context: 256K, License: other.

8-bit Base model:moonshotai/kimi-k2.... Base model:quantized:moonshota... Blackwell Compressed-tensors Conversational Custom code Image-text-to-text Laguna Laguna-xs-2.1 Moonvit Multimodal Nvfp4 Region:us Safetensors Sharded Tensorflow Vision Vision-language

Model Card on HF 🤗: https://huggingface.co/webbrain-one/Laguna-XS-2.1-Vision-NVFP4

Laguna XS 2.1 Vision NVFP4 Parameters and Internals

LLM Name	Laguna XS 2.1 Vision NVFP4
Repository 🤗	https://huggingface.co/webbrain-one/Laguna-XS-2.1-Vision-NVFP4
Base Model(s)	poolside/Laguna-XS-2.1-NVFP4 moonshotai/Kimi-K2.6 poolside/Laguna-XS-2.1-NVFP4 moonshotai/Kimi-K2.6
Model Size	33.4b
Required VRAM	21.5 GB
Updated	2026-08-09
Maintainer	webbrain-one
Model Type	laguna
Model Files	0.1 GB 5.1 GB: 1-of-5 5.1 GB: 2-of-5 5.1 GB: 3-of-5 5.1 GB: 4-of-5 1.1 GB: 5-of-5 0.8 GB
Model Architecture	LagunaForConditionalGeneration
License	other
Context Length	262144
Model Max Length	262144
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	〈\|PAD\|〉
Vocabulary Size	100352
Torch Data Type	bfloat16

Rank the Laguna XS 2.1 Vision NVFP4 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