moondream

ComfyUI nodes for vision-language models: Qwen3-VL, Moondream 3, Florence-2, SmolVLM2, InternVL, Gemma 3, MiniCPM-V. Plus open-vocabulary detection, SAM2/SAM3 segmentation, video temporal reasoning, GGUF via llama.cpp, and hosted LLM/VLM APIs.

AI-powered assistant to help you with your daily tasks, powered by Llama 3, DeepSeek R1, and many more models on HuggingFace.

A CPU Realtime VLM in 500M. Surpassed Moondream2 and SmolVLM. Training from scratch with ease.

AI-powered visual reasoning tools for broadcast & ProAV. PTZ camera tracking, object detection, scene analysis using Moondream VLM. By StreamGeeks & PTZOptics.

The vision plugin for OpenCode that truly understands images. Inspect, read, and reason about any screenshot or picture with deeper understanding than any other plugin — fully local, private, and free.

tiny vision language model

ComfyUI wrapper for Moondream's gaze detection

Use the Moondream 2 model to detect faces and their gaze directions in videos.

Moondream MCP Server in Python

Python scripts to use for captioning images with VLMs

Moondream is a lightweight multimodal large language model

A powerful video summarization tool that utilizes Moondream alongside multiple AI models to provide comprehensive video understanding through audio transcription, intelligent frame selection, visual description, and content summarization.

Automatically annotates YOLO dataset using Moondream visual model

An open-source AI tool that analyzes images (from file or URL) and generates detailed descriptions. Perfect for Computer Vision projects, alt-text automation, and learning. Includes a ready-to-use API.

🐦 OpenClaw skill — Live bird detection with YOLOv11 + Moondream VLM species ID. Real-time annotated video stream to any device.

Modal moondream inference

Welcome to Malaika my ai companion

full text search your meme/screenshot folder with small LMs x traditional OCR

This tool uses Moondream 2B, a powerful yet lightweight vision-language model, to detect and redact objects from videos. Moondream can recognize a wide variety of objects, people, text, and more with high accuracy while being much smaller than most vision models.

Integrate moondream2 with the Telegram api