PhotoPrism 260523-0544f71c1

This update introduces a redesigned Info Sidebar that lets you edit metadata, albums, and labels as well as manually tag faces without leaving the full-screen viewer. On the AI side, our ONNX-based face recognition pipeline has fully replaced the legacy Pigo detector, and the vision.yml configuration now accepts mixed-case model names so all identifiers from Hugging Face, Ollama, and OpenAI-compatible catalogs can be used.

Media handling has been thoroughly modernized: video transcoding now supports Vulkan hardware acceleration via FFmpeg 8, images use a native HEIC/AVIF reader (with libheif upgraded to v1.21.2), and layered TIFF and Adobe Photoshop PSD files are now supported. Other highlights include NOT & AND operators in the label filter, a drag-and-drop file upload zone, zstd compression for faster page loads, hardened WebDAV interoperability, and a new Ubuntu 26.04 LTS base image.

A big thank you to everyone who contributed and helped with testing! We hope you enjoy this release. 🌈💎✨

What's new?

• Viewer: Sidebar shows editable metadata, albums, and labels by @omerdduran
• Viewer: Captions can be hidden using the menu or a keyboard shortcut
• Faces: Viewer sidebar allows to manually tag faces on pictures by @omerdduran
• Faces: Dropped legacy Pigo detector in favor of ONNX-based detection
• UX: Added a drag-and-drop zone to the file upload dialog
• UX: Improved form input validation and numeric range caps
• Login: Enhanced login page with "Stay signed in on this device" toggle
• Login: OIDC provider initialization is retried after transient discovery failure
• Login: Fixed OIDC redirect of unauthenticated users when opening direct links
• Search: Improved label filter with support for NOT & AND operators
• Labels: Added support for homophones and homophone-aware lookups by @keif888
• Labels: Fixed duplicates when renaming a label and re-adding the previous name
• Labels: Fixed case and punctuation variants creating duplicates in the edit dialog
• Folders: Child folders with emoji paths no longer overwrite parent albums
• Videos: Added support for Vulkan hardware transcoding using FFmpeg 8
• Videos: HEVC remux output is tagged as hvc1 based on MP4 chunk scan
• Images: Added support for layered TIFF and Adobe Photoshop PSD images
• Images: Added native HEIC/AVIF reader and upgraded libheif to v1.21.2
• Images: Replaced disintegration/imaging library with native format support
• Server: Added zstd compression support for faster loading times
• Server: Added pre-compressed frontend bundles for faster loading times
• Server: Fixed nil-DB race in async count and cover update goroutines
• WebDAV: Hardened timeouts, cancellation, and Depth-1 fallback diagnostics
• WebDAV: Fixed settings dialog to allow credentials for existing services to be changed
• WebDAV: Added fallback for servers that only allow PROPFIND with a Depth of 1
• CLI: Improved vision run command updates sidecar YAML files
• CLI: Added a faces config subcommand to list face-related options
• CLI: Fixed flags placed after positional arguments being silently dropped
• MCP: Added read-only support for the Model Context Protocol (MCP)
• MCP: Added --disable-mcp flag to disable Model Context Protocol support
• Config: Removed limitation for vision model names to be lowercased
• Config: Improved worker auto-configuration based on number of CPU cores
• Config: Consolidated SQL driver names and parsing in pkg/dsn by @keif888
• Config: Default HTTP and HTTPS ports are stripped from base URLs
• Logs: Information about long-running processes is logged by @keif888
• Docker: Upgraded base image to Ubuntu 26.04 LTS (Resolute Raccoon)
• Security: Search queries now use parameterized statements for all user input
• Security: Upgraded Go from v1.26 to v1.26.3 and ONNX Runtime to v1.25.1

Redesigned Info Sidebar

Edit a metadata field by clicking on it. Some fields can be edited in place, while others open a dialog box.

Editing People & Faces

The viewer sidebar offers the same face management actions that are already available under Edit > People. In addition, it allows you to manually mark and assign faces that were missed by the automatic detection.

Drag-and-Drop File Upload

Drop any number of files or folders onto the upload area, or click it to open a file picker. Files selected for upload are listed with their file names, sizes, and a thumbnail, if possible.

Translations

Missing user interface translations have been generated with the help of DeepL and Google Translate. Native speakers are welcome to help us improve them where needed. Learn more ›

Installation Packages

The packages attached to this release are intended for experienced users and maintainers of third-party integrations only, as they require manual configuration and do not include tested system dependencies. Since we are unable to provide support for custom installations, we recommend using one of our Docker images to run PhotoPrism on a private server or NAS device. Learn more ›