Label Studio

Health v2 · maintenance only

Adoption not part of the score

28150 stars · 3678 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Label Studio is an open-source data labeling and annotation platform supporting images, audio, text, video, and time series with a web UI and standardized export formats. It includes an ML backend for model-assisted pre-labeling, a Python SDK, API, and integrations with cloud storage and foundation models.

Use cases

• label images for object detection training
• annotate text for named entity recognition
• create semantic segmentation datasets
• label audio for speech recognition
• collect human preference data for RLHF
• evaluate LLM outputs with human review
• pre-label images with YOLO or SAM models

When to choose

• you need a self-hosted multi-modality annotation tool
• you want ML-assisted pre-labeling in a human-in-the-loop workflow
• you need standardized export formats for training pipelines
• you want API/SDK automation of labeling projects

When to avoid

• you need only tiny one-off labeling done in a spreadsheet
• you require the enterprise features like advanced review queues without a paid plan
• you want a fully offline desktop tool with no server

Facets

application · maturity active

machine-learning data-science image-processing nlp audio-processing video-processing sdk self-hosted machine-learning computer-vision data-science artificial-intelligence developer-tools python self-hosted cross-platform cloud data-labeling annotation human-in-the-loop mlops dataset-preparation computer-vision bounding-boxes semantic-segmentation rlhf llm-evaluation natural-language-processing web-server docker kubernetes

Member repositories

For agents

markdown · JSON · MCP: product_card(name="HumanSignal/label-studio")