For the complete documentation index, see llms.txt. This page is also available as Markdown.

2024 RELEASES

Unitlab product updates released in 2024.

v1.1.7 — Bounding-box exports, billing, and account synchronization

Features

  • Added billing-page access, dynamic class ordering, bounding boxes across render and COCO/YOLO export paths, and optional empty releases.

  • Added account and subscription synchronization with Customer.io, plus path-based authentication support.

Fixes

  • Corrected thumbnail annotations, subscription updates, timestamp conversion, permissions, model filtering, and environment-specific Sentry behavior.

v1.1.6 — Upload quota enforcement and security patches

Features

  • Added subscription-quota checks to project and SDK data uploads, including concurrent-upload enforcement.

Fixes

  • Upgraded Django to address security advisories.

v1.1.5 — Existing-project uploads and user-provided LLM annotation

Features

  • Added data upload into existing projects.

  • Added an OpenAI-based LLM auto-annotator using credentials supplied by the workspace.

  • Expanded release identifiers, administrative reporting, and AI integration utilities.

Fixes

  • Corrected reviewer statistics and assignment after new uploads, free-plan quota behavior, missing-file handling, and subscription-state checks.

  • Updated CORS, certificates, and security-sensitive dependencies.

v1.1.4 — Multi-owner workspaces and dynamic subscriptions

Features

  • Added multiple workspace owners, Free and Premium subscription handling, dynamic pricing tables, and customer-specific plan information.

  • Expanded member restrictions, invitation status, and subscription-aware quick-start actions.

Fixes

  • Removed the single-owner constraint and corrected permission, email, and Free-plan migration behavior.

v1.1.3 — Onboarding, quick demos, and admin analytics

Features

  • Added database-synchronized onboarding and quick-start actions, one-click demo project creation, generic data types, and expanded admin statistics and CSV exports.

  • Added invitation email updates and file-checksum tooling.

Fixes

  • Corrected annotation-tool and SDK behavior, COCO output, documentation links in email, distinct-value filtering, and subscription cancellation checks.

v1.1.2 — Instance segmentation and large-project performance

Features

  • Added instance-segmentation annotation, COCO/RLE import and export, SDK post-processing, background deletion, and digest-aware dataset updates.

  • Added precalculated project, dataset, annotator, and reviewer counts for faster lists and progress reporting.

Fixes

  • Reworked project and release deletion, dataset cloning, storage I/O, temporary-file cleanup, and EFS locking for better reliability.

  • Corrected permissions, distinct filters, pending invitations, and result ordering.

v1.1.1 — Expanded export formats and independent releases

Features

  • Added YOLOv5, YOLOv8, and COCO conversion for boxes, polygons, segmentation, skeletons, lines, and points.

  • Made releases independent from their source project, added dynamic ZIP delivery and thumbnail generation, and expanded SDK dataset create/upload/download workflows.

Fixes

  • Improved safe filenames, temporary storage, result preparation, area sorting, dataset-list queries, and request throughput.

  • Removed duplicate CLI code by combining command-line behavior with the SDK and removed redundant application throttling behind Cloudflare.

v1.1.0 — Licenses, demo releases, and export conversion

Features

  • Added license lists, demo releases, editable AI models, SDK dataset downloads, dataset rotation metadata, and YOLOv5 export.

  • Allowed managers to manage project classes and exposed more release and datasource progress information.

Fixes

  • Removed list bottlenecks, improved pagination and ordering, corrected cloned-project defaults, and enforced unique grayscale class colors.

  • Reworked export conversion to use a more memory-safe implementation.


Related Unitlab capability guides: AI training-data annotation · image annotation for computer vision