What are the hardware requirements for Moondream2?

Moondream2 requires approximately 3.9 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Moondream2 and how large is it?

Moondream2 is developed by vikhyatk, a model with 1.9b 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 Moondream2?

Moondream2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Moondream2 by vikhyatk — VRAM 3.9GB

Name: Moondream2
Author: vikhyatk

Moondream2 is an open-source language model by vikhyatk. Features: 1.9b LLM, VRAM: 3.9GB, License: apache-2.0, LLM Explorer Score: 0.25.

Custom code Doi:10.57967/hf/6762 Endpoints compatible Eval-results Image-text-to-text Moondream1 Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/vikhyatk/moondream2

Moondream2 Parameters and Internals

LLM Name	Moondream2
Repository 🤗	https://huggingface.co/vikhyatk/moondream2
Model Size	1.9b
Required VRAM	3.9 GB
Updated	2026-07-09
Maintainer	vikhyatk
Model Type	moondream1
Model Files	3.9 GB
Model Architecture	HfMoondream
License	apache-2.0
Model Max Length	2048
Transformers Version	4.52.4
Tokenizer Class	CodeGenTokenizer
Torch Data Type	bfloat16

Rank the Moondream2 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