A practical list of AI art galleries, prompt libraries, datasets, model hubs, and open tools for exploring generated images.
This is intentionally focused on useful browsing, prompt study, provenance, and open metadata rather than hype.
- Generated Gallery - free AI art gallery with prompts, source links, tags, categories, and SFW/NSFW filtering.
- Lexica - Stable Diffusion image and prompt search.
- Krea - AI image generation and visual inspiration boards.
- OpenArt - AI image generation, discovery, and workflows.
- Civitai Images - large community gallery for model outputs and prompts.
- PromptHero - prompt examples across image generation models.
- Generated Media Protocol - JSONL schema for AI-generated images, prompts, provenance, safety labels, and portable gallery packs.
- Generated Gallery AI Image Dataset - public metadata feed with manifest, sample records, JSON schema, prompts, and source links.
- LAION - open datasets and research around image-text data.
- DiffusionDB - prompt and generated image dataset for Stable Diffusion research.
- Hugging Face Models - text-to-image and image generation models.
- Civitai Models - community model, LoRA, and checkpoint hub.
- Replicate Explore - hosted AI model demos and APIs.
- Prompt Engineering Guide - broad prompt engineering reference.
- Stable Diffusion Web UI - popular local Stable Diffusion interface.
- ComfyUI - node-based image generation workflows.
- InvokeAI - professional creative interface for Stable Diffusion.
- Machine Dream Finds - small themed AI image packs with prompt fragments, source trails, safety labels, and manifests.
- Haunted Product Catalog - a Machine Dream Finds example pack with luxury objects, cursed studio lighting, prompts, and source trails.
- Generated Gallery Creator Kit - guide for publishing AI image packs that another viewer or agent can inspect.
- Protocol docs - portable generated-media record format.
- Creator kit - how to publish small AI image packs.
- Manifest - current feed metadata, checksum, record count, and schema links.
- SFW JSONL export - compressed public metadata feed.
- Prompt-only export - prompt and label focused split for lightweight analysis.
Good resources preserve at least some of the context around generated media:
- prompt text or fragments
- model or workflow metadata
- source URLs and provenance
- tags or categories
- safety labels
- exportable feeds, schemas, APIs, or open repos
PRs welcome. Please add resources that are useful for AI art discovery, prompt research, open metadata, model exploration, or reproducible image workflows.
Avoid low-quality tool farms, affiliate-only pages, and sites that hide all useful context behind a signup wall.