SnapOtter

Version 2.2.0 focuses on security hardening, patching several privilege escalation and access control vulnerabilities, while adding new image tools and improving accessibility and self-hosting stability.

This hotfix release includes numerous bug fixes for the demo and landing pages, improved AI feature-bundle installations, and updates to the navigation and file-processing infrastructure.

SnapOtter 2.0 expands into a full file-manipulation suite with 200+ tools across image, video, audio, and PDF modalities, featuring a new layer-based editor and Redis-backed background jobs.

Introduces an HTML to Image tool, WCAG 2.2 AA accessibility compliance, security hardening, and various Docker and AI tool fixes.

Introduces a live demo, tool index, 52 SEO landing pages, background previews, and various bug fixes and infrastructure improvements.

Introduces a full image editor, SSO authentication, five new tools, support for 20 languages, and expanded file format compatibility.

Version 1.16.0 introduces several improvements and updates to the SnapOtter platform.

This release includes minor bug fixes and stability improvements.

This release includes bug fixes and minor improvements.

Introduces an Automate page redesign, extended image format support including AVIF, production-grade RBAC, and various security and stability fixes.

This release fixes a bug by changing how Node.js is copied into the image, using the official image instead of apt-get.

Fixed a bug by adding retry with backoff for apt-get update on the CUDA base image.

Parallelized model downloads and implemented a registry cache for improved performance.

Excludes e2e-docker tests from the Vitest runner.

This release fixes issues related to verbose error handling, batch processing, and multi-file support.

Fixed an issue by adding retries to apt-get update during transient mirror sync errors.

Fixed a bug to ensure GHCR login uses GHCR_TOKEN with the write:packages scope.

This release includes several bug fixes related to Docker image builds and runtime model path resolution for non-root users.

This release introduces a unified Docker image with GPU auto-detection, expanded model pre-downloading, and numerous bug fixes for PaddleOCR, RealESRGAN, and arm64 compatibility.

This release implements a comprehensive RBAC system with permission maps, new middleware for route protection, and extended authentication hooks.

Introduces a new pdf-to-image tool with frontend and backend support and a unified Docker image featuring GPU auto-detection.

Introduces a new stitch tool featuring an API route handler, settings UI component, and integration into the tool registry and i18n.

Added variant diagnostics to the health endpoint and lite mode banner.

Added automatic generation of llms.txt using the vitepress-plugin-llms plugin.

Introduces content-aware resizing features via a new seam carving AI bridge, Python scripts with face protection, and updated UI settings.

Introduces metadata editing capabilities including a new API route, UI components, and server-side filtering for unsafe keys.

Improved the reliability of AI tools for face detection and background removal.

Fixed issues regarding batch SSE progress and non-AI processing user experience.

Added server-side logging to AI tool routes.

Fixed issues with batch file ordering and format preservation for image tools.

Fixed broken llms.txt links on the REST API documentation page.

Fixed GPU detection by using torch.cuda instead of onnxruntime providers.

Fixed a bug where GPU detection occurred incorrectly in CUDA images running without a GPU.

Introduces GPU/CUDA acceleration support and fixes an issue with empty Authorization headers for forward-auth proxy compatibility.

Introduced a lightweight Docker image without AI/ML tools via the :lite tag.

This release fixes HEIC decoding on Linux by utilizing heif-convert.

Fixed a bug by installing HEVC codec plugins for CI HEIC tests.

Fixed an issue by installing libheif-examples in CI for HEIC tests.

Added support for HEIC and HEIF formats for both input and output.

Added a 'Crop to Content' mode to the smart crop tool.

This release fixes a bug by setting the default theme to light instead of following the system preference.

This release introduces new utility functions for clipboard copying and ID generation to improve compatibility and fixes various bugs related to XHR timeouts, Docker processing, and volume permissions.

Fixed volume permission issues when using a bind-mounted /data directory.

Relocates the theme toggle and GitHub button to the top-right navigation bar.

Added a GitHub stars button to the documentation navbar and fixed the footer license.

This release adds support for SVG files in the convert tool.

Major 1.0.0 release introducing a full React SPA, Fastify API server, SQLite database, comprehensive AI and image processing tools, authentication, and batch processing capabilities.