English | 中文

👉Official Site

TIOL — AI Local Photo Manager

Your privacy-first, local-first photo manager. Automatically tags photos. Built-in fast, intelligent searching.

Supports MacOS, Windows, and Linux.

All photos stay on your own hard drive. AI inference runs completely offline. No cloud services required.

Features

• 🔍 AI Powered Dual Search: Semantic search (simply describe what you're looking for, e.g., "a cup of coffee") + Tag search (fast)

• 📄 Filter redundant images: One-click organization of redundant and blurry/overexposed images.

• 📚 Albums: user albums plus lens & color smart groups; multi-select import and drag-and-drop.

• 💻 Multiplatform support: Supports Windows, MacOS, and Linux. Supports Nvidia CUDA, cpu, Apple CoreML，etc.

• ⚡ Lightweight: ~50MB in size，single exe and system onnx driver，provides portable editions。

• 📷 Lens/Focal length filtering: Filter the photos through specifying the lens/the focal length.

• 📁 Directory Management: Add/remove photo directories, file system monitoring (new/modified files are automatically queued for processing)

• 🏷️ Custom Labels: AI tagging – enter any labels (Chinese or English) in the Tags tab; files are indexed automatically on changes.

• 🖥️ AI Engine Options: auto / GPU / CPU / Apple CoreML (macOS native Neural Engine acceleration) – fully local

• 🔒 Model Lockdown: SHA256 checksum + resumable downloads + fallback to domestic mirrors, auto‑repair for corrupted models

Installation

Compiled Installers & Portable editions Official Website (With Compiled Releases)

Quick Start

# Dependencies: Rust 1.70+, tauri-cli 2.x (macOS also needs Xcode CLT; Windows needs VS Build Tools)
cargo tauri dev        # development mode
cargo tauri build      # production build (Windows → msi/nsis, macOS → .app/dmg)

On first launch, the AI model (~412 MB) will be downloaded automatically (mirror chain: hf‑mirror → openi → huggingface).
Before building/running on Windows, copy vendor/onnxruntime/win-x64/onnxruntime.dll to the same directory as the executable.
On macOS, you need a universal2 onnxruntime.dylib (official builds include CoreML EP) – see BUILD.md for details.

Data Locations

• Database / thumbnails:
  Windows: %APPDATA%\com.tiol.desktop
  macOS: ~/Library/Application Support/com.tiol.desktop
• AI models:
  Windows: %LOCALAPPDATA%\com.tiol.desktop\models
  macOS: ~/Library/Caches/com.tiol.desktop/models

Contributing

This project is iterating quickly; issues and pull requests (as long as they are reasonable) are welcomed.

Special Thanks

Thanks to Ken709-mp4 for providing image examples,patches, and ideas.

Thanks to DiegoTang for providing image examples and ideas.

👉Link To Their Site