<p align="center">
<a href="https://lightly.ai/lightly-studio">
<picture>
<source
media="(prefers-color-scheme: dark)"
srcset="https://storage.googleapis.com/lightly-public/studio/lightlystudio_standard_horizontal_light.png"
/>
<source
media="(prefers-color-scheme: light)"
srcset="https://storage.googleapis.com/lightly-public/studio/lightlystudio_standard_horizontal_dark.png"
/>
<img
src="https://storage.googleapis.com/lightly-public/studio/lightlystudio_standard_horizontal_dark.png"
height="50"
alt="LightlyStudio logo"
/>
</picture>
</a>
</p>
<p align="center"><strong>Curate, Annotate, and Manage Your Data in LightlyStudio.</strong></p>
<p align="center">
<a href="https://pypi.org/project/lightly-studio"><img src="https://img.shields.io/pypi/pyversions/lightly-studio" alt="PyPI python" /></a>
<a href="https://pypi.org/project/lightly-studio"><img src="https://badge.fury.io/py/lightly-studio.svg" alt="PyPI version" /></a>
<a href="LICENSE"><img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg" alt="License" /></a>
<a href="https://docs.lightly.ai/studio"><img src="https://img.shields.io/badge/Docs-blue" alt="Docs" /></a>
<a href="https://colab.research.google.com/github/lightly-ai/lightly-studio/blob/main/lightly_studio/src/lightly_studio/examples/example_notebook.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab" /></a>
</p>
---
<p align="center">
<img src="assets/readme/hero_showcase.gif" alt="LightlyStudio: an image grid with annotations next to an embedding plot, a cluster selection filtering the grid, and the annotation editor open on one image" width="100%" />
</p>
<p align="center">The embedding plot shows how images relate to each other, with a preview on hover. A lasso selection filters the grid to one cluster. A search for "coffee" finds a match, and the annotation editor opens to label it.</p>
<p align="center"><strong>⚡ Works smoothly with 2M+ images, embeddings included, on a single MacBook (M1, 16GB RAM).</strong></p>
## 🚀 Try it in 60 seconds
LightlyStudio runs on your computer and opens in your browser. The following command downloads an example dataset with images, annotations, and evaluation results — no account needed.
```bash
pip install lightly-studio
lightly-studio quickstart
```
- **Local only:** your images and datasets never leave your machine.
- **Your own images:** one Python call indexes them, then start the server. See the [Image Dataset guide](https://docs.lightly.ai/studio/workflows/image_dataset/).
- **Your own videos:** see the [Video Dataset guide](https://docs.lightly.ai/studio/workflows/video_dataset/).
- **No install:** [open the quickstart in Colab](https://colab.research.google.com/github/lightly-ai/lightly-studio/blob/main/lightly_studio/src/lightly_studio/examples/example_notebook.ipynb)
- **A guided walkthrough:** read the [Tutorials](#-tutorials) below. [Curate a Traffic CCTV Dataset for YOLO Training](https://docs.lightly.ai/studio/tutorials/yolo-traffic-cctv-object-detection/) goes from raw images to a trained model.
- Runs on **Python 3.10 to 3.14** on Windows, Linux, and macOS. Use Python 3.10 for plugin compatibility, for example [SAM autolabeling](https://docs.lightly.ai/studio/ecosystem/plugins/#example-lightlystudio-sam3-plugin).
## Workflows
<table>
<tr>
<td align="center">
<a href="https://docs.lightly.ai/studio/workflows/image_dataset/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/image_dataset.png" width="400" alt="Image Datasets"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/workflows/image_dataset/">Image Dataset</a></strong>
</td>
<td align="center">
<a href="https://docs.lightly.ai/studio/workflows/video_dataset/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/video_dataset.png" width="400" alt="Video Dataset"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/workflows/video_dataset/">Video Dataset</a></strong>
</td>
<td align="center">
<a href="https://docs.lightly.ai/studio/workflows/annotations/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/annotation.png" width="400" alt="Annotate"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/workflows/annotations/">Annotation</a></strong>
</td>
</tr>
<tr>
<td align="center">
<a href="https://docs.lightly.ai/studio/workflows/sampling/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/sampling.png" width="400" alt="Curate"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/workflows/sampling/">Curation</a></strong>
</td>
<td align="center">
<a href="https://docs.lightly.ai/studio/ecosystem/plugins/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/plugins.png" width="400" alt="Plugins"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/ecosystem/plugins/">Plugins</a></strong>
</td>
<td align="center">
<a href="https://docs.lightly.ai/studio/workflows/evaluation/">
<img src="https://storage.googleapis.com/lightly-public/studio/docs_cards/model_evaluation.png" width="400" alt="Model Evaluation"/>
</a>
<br/><strong><a href="https://docs.lightly.ai/studio/workflows/evaluation/">Model Evaluation</a></strong>
</td>
</tr>
</table>
## Example workflows
LightlyStudio is a browser app that runs on your own computer. Load your data into the local
database with a Python script, then start the server and explore it in your browser.
<details open>
<summary><strong>Index a COCO dataset</strong></summary>
Create a file named `example_coco.py`:
```python
import lightly_studio as ls
# Download the example dataset (will be skipped if it already exists)
dataset_path = ls.utils.download_example_dataset(download_dir="dataset_examples")
dataset = ls.ImageDataset.load_or_create()
dataset.add_samples_from_coco(
annotations_json=f"{dataset_path}/coco_subset_128_images/instances_train2017.json",
images_path=f"{dataset_path}/coco_subset_128_images/images",
)
# Start the UI server on localhost:8001.
# Pass `host` and `port` parameters to customize it.
# Reopen the app later with `lightly-studio gui` instead of re-running this script.
ls.start_gui()
```
Run `python example_coco.py` and open the printed URL to inspect images with their annotations.
To ingest COCO samples with segmentation masks instead of object detections, pass
`annotation_type=ls.AnnotationType.SEGMENTATION_MASK` to `add_samples_from_coco()`.
</details>
<details>
<summary><strong>Index a folder of images</strong></summary>
Create a file named `example_image.py`:
```python
import lightly_studio as ls
# Download the example dataset (will be skipped if it already exists)
dataset_path = ls.utils.download_example_dataset(download_dir="dataset_examples")
# Index the images, create embeddings, and store everything in the local database.
dataset = ls.ImageDataset.load_or_create()
dataset.add_images_from_path(
path=f"{dataset_path}/coco_subset_128_images/images",
)
# Start the UI server on localhost:8001.
# Pass `host` and `port` parameters to customize it.
ls.start_gui()
```
Run `python example_image.py` and open the printed URL in your browser.
</details>
Other starting points:
[YOLO datasets](https://docs.lightly.ai/studio/workflows/image_dataset/),
[video folders](https://docs.lightly.ai/studio/workflows/video_dataset/),
[custom annotations](https://docs.lightly.ai/studio/workflows/annotations/#annotations-in-python),
[model evaluation](https://docs.lightly.ai/studio/workflows/evaluation/#model-evaluation-in-python),
[notebooks](https://docs.lightly.ai/studio/get_started/notebooks/),
and more in the [docs](https://docs.lightly.ai/studio/).
## 📚 Tutorials
Step-by-step guides covering complete workflows — from raw, unlabeled data to a trained and evaluated model:
- **[Curate a Traffic CCTV Dataset for YOLO Training](https://docs.lightly.ai/studio/tutorials/yolo-traffic-cctv-object-detection/):** Explore embeddings, remove near-duplicates, auto-label with a detection plugin, and review annotations before training.
- **[Evaluate YOLO26 on Your Dataset with LightlyStudio](https://docs.lightly.ai/studio/tutorials/yolo26-model-evaluation/):** Compare predictions against ground truth, use the confusion matrix and embeddings to find failure patterns, and export issues for relabeling.
## A selection of features
**Set up a dataset**
- [Image datasets](https://docs.lightly.ai/studio/workflows/image_dataset/) — index folders, COCO, and YOLO data
- [Video datasets](https://docs.lightly.ai/studio/workflows/video_dataset/) — index videos and work with frames
- [Cloud storage](https://docs.lightly.ai/studio/ecosystem/cloud_storage/) — read from S3, GCS, or Azure
- [Reuse datasets](https://docs.lightly.ai/studio/get_started/reuse_datasets/) — append to an existing database, or point at a custom path
- [Notebooks](https://docs.lightly.ai/studio/get_started/notebooks/) — run the GUI inside Jupyter or Colab
**Work with your data**
- [Query, filter, and sort](https://docs.lightly.ai/studio/workflows/search_and_filter/#query-in-python) — build reusable subsets in Python
- [Lightly Query Language](https://docs.lightly.ai/studio/workflows/lightly_query_language/) — the query syntax for the GUI query editor
- [Annotations](https://docs.lightly.ai/studio/workflows/annotations/), [Tags](https://docs.lightly.ai/studio/workflows/tags/), [Metadata](https://docs.lightly.ai/studio/workflows/metadata/), [Captions](https://docs.lightly.ai/studio/workflows/captions/), [Embeddings](https://docs.lightly.ai/studio/workflows/embeddings/)
- [Sampling](https://docs.lightly.ai/studio/workflows/sampling/) — pick the most typical and diverse samples to cut labeling cost
- [Model evaluation](https://docs.lightly.ai/studio/workflows/evaluation/) — compare predictions against ground truth
- [Plugins](https://docs.lightly.ai/studio/ecosystem/plugins/) — auto-label with SAM and other models
- [Export](https://docs.lightly.ai/studio/workflows/export/) — write a query back out as COCO or YOLO
## 🐍 Python Interface
Everything in LightlyStudio is scriptable: index datasets, query and edit samples, sample
subsets, and export the result. See the
[API reference](https://docs.lightly.ai/studio/api/dataset/).
## 🤝 Contribute
We welcome contributions! Please check our issues page for current tasks and improvements, or
propose new issues yourself. See [CONTRIBUTING.md](CONTRIBUTING.md) for development setup and
validation instructions.
## 💬 Contact
[](https://www.lightly.ai/lightly-studio) <br>
[](https://discord.gg/xvNJW94) <br>
[](https://github.com/lightly-ai/lightly-studio) <br>
[](https://x.com/lightlyai) <br>
[](https://www.youtube.com/channel/UCAz60UdQ9Q3jPqqZi-6bmXw) <br>
[](https://www.linkedin.com/company/lightly-tech)