Alibaba Cloud PAI iTAG vs BRAT

Alibaba Cloud PAI iTAG is a data-labeling platform for image, text, video, audio and multimodal datasets. Built-in image tasks include classification, object detection, OCR, table recognition and semantic segmentation; official templates also cover image-text explanation, conversation ranking, visual question answering and video OCR. Teams can use official or tenant templates, or create templates for business-specific labeling. A job brings together an OSS-backed dataset and template with labels, subtasks, checks, labelers and reviewers. Manual labeling is free, and selected large-model labeling templates are currently free; OSS storage and data transfer are billed separately. Input data must be in Alibaba Cloud OSS, and the workspace and bucket must be in the same region. Dataset imports require an index file rather than a direct path to a folder of raw files. From August 1, 2026, iTAG 1.0 is allowlist-only and in maintenance mode. Existing customers with prior labeling tasks or datasets can continue using current features. New users seeking AI-powered automatic dataset labeling are directed to Multimodal Data Management. API operations include listing, creating, updating and deleting templates and labeling tasks.

brat is a free, open-source tool for structured text annotation, with annotations represented in forms computers can process. It supports text spans, binary relations, n-ary associations, attributes, normalization annotations, and freeform notes. Teams can configure categories, types, and constraints for their tasks, and multiple annotators can work on the same document or collection while seeing one another’s edits nearly as they happen. Normalization can link annotations to external resources, and brat can incorporate output from automatic annotation services available as web services. Its server and client support Unicode across nearly 100 scripts. The tool checks edits against configured constraints and searches documents or collections, including keyword-in-context concordancing. Annotations export in brat stand-off format; visualizations can be saved as SVG, PNG, PDF, or EPS. The free v1.3 release is self-hosted under the MIT License. New server installations assume a UNIX-like environment; Windows users can run brat in a UNIX-like virtual machine. The experimental standalone server is not security-hardened and is not recommended for sensitive data or systems accessible from the internet.