#3131st of 34 in Open models
moondream
Small vision language model for captioning, visual Q&A and object detection
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 10.1k
- License
- Apache-2.0
- Last commit
- Apr 2026
- Language
- Python
9,438 on 31 Mar 202610,128 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Moondream is an open-weight vision language model that answers questions about images, writes captions and detects objects. It comes in two sizes: a 2B-parameter general model and a 0.5B model meant as a distillation target for edge devices. It can be run locally or through a hosted cloud service, with setup covered in the external quickstart docs.
Who it is for: Developers adding image understanding to apps, locally or on edge devices
Strengths
- Two sizes: 2B for general use, 0.5B for resource-constrained edge devices
- Covers captioning, visual question answering and object detection in one model
- Apache-2.0 license, with a hosted playground for trying it before installing
Weaknesses
- README gives no RAM, VRAM or latency figures
- README has no install or run steps; it points to external docs
- No Docker setup or tagged release in the repository
- No benchmark results in the README
What it needs
- Models: Moondream 2B, Moondream 0.5B
Also in Open models
See all 34| Rank | Project | Score |
|---|---|---|
| 1 | llmfitRust tool that checks which LLMs fit your RAM, CPU and GPU | 87 out of 100 |
| 2 | layaSingle-pass multilingual classifier for typed choice, score and yes/no decisions | 84 out of 100 |
| 3 | timesfmPretrained time-series foundation model for zero-shot forecasting from Google Research | 75 out of 100 |
| #4 | speechPyTorch toolkit for training and deploying ASR, TTS and speech LLM models | 70 out of 100 |
| #5 | yueLyrics-to-song model that writes an editable melody and chord score first | 65 out of 100 |