Label Studio

Health v2 · maintenance only

Adoption not part of the score

28488 stars · 3754 forks observed · 2026-10-19

What it is AI-extracted, prompt v1, taxonomy v1, 2026-10-05, confidence 0.99

Label Studio is an open‑source, web‑based data labeling and annotation platform that supports text, image, audio, video, and time‑series data. It provides a flexible UI, API, and SDKs to create custom labeling workflows and integrate machine‑learning models.

Use cases

• Create bounding‑box annotations for object detection datasets
• Label audio recordings for speech‑to‑text model training
• Annotate text documents for named‑entity recognition
• Generate segmentation masks for image segmentation projects
• Collect human feedback for LLM fine‑tuning and RLHF
• Prepare multi‑modal datasets (image + text) for multimodal AI models

When to choose

• You need a single tool that can label many data types (text, image, audio, video, time‑series).
• You want a self‑hosted solution with a customizable UI and API for integration into ML pipelines.
• You require human‑in‑the‑loop workflows with model pre‑annotations and active learning.

When to avoid

• You only need a lightweight command‑line labeling script without a web UI.
• Your organization requires a fully managed SaaS solution with built‑in compliance certifications.
• You need real‑time collaborative labeling with fine‑grained permission controls not yet supported.

Facets

application · maturity active

image-processing video-processing nlp web-framework api-framework cli data-science computer-vision windows browser cli data-labeling human-in-the-loop multi-modal open-source sdk ml-integration custom-templates annotation natural-language-processing audio-processing video mlops linux macos docker web-server

Member repositories

For agents

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