LLMII

LLMII labels image collections with AI-generated captions and keywords for people who want searchable photos without uploading them to a cloud service. It processes folders on your computer and saves descriptions that other photo tools can read.

• Portable metadata: Write captions and keywords into image files or separate XMP sidecars, so the labels stay usable in software that reads standard metadata fields.
• Collection processing: Process images across subfolders, including major camera RAW formats, and resume interrupted work without repeating completed files.
• Caption and keyword control: Choose simple or detailed captions, preserve existing keywords, or add labels from another model without duplicating keywords.
• Visual review: Step through processed images in the desktop interface to check the generated descriptions and tags.

LLMII is open source under GPL-3.0 and runs on Windows, Linux, and Apple Silicon Macs. It uses KoboldCpp for local AI processing. After the models and dependencies have downloaded, all components work locally and offline, with no cloud service or database required.

GPU acceleration supports Apple Metal, NVIDIA CUDA, and AMD through Vulkan; a discrete GPU is recommended. The tool can modify image metadata directly, while its XMP sidecar option leaves image files unchanged.