Image explorer — five dimensions of visual discovery
Endlss is an image exploration tool that connects images across visual similarity, colour, tags, prompts, and names. Select an image, see what's similar, select again — every click reveals something new.
How it works
Select an image. See what's connected. Follow the thread.
Traditional image search starts with words. You type a query, get results, and refine your terms until you find what you're looking for. But some of the best visual discoveries happen when you don't know what you're looking for.
Endlss flips the model. Start with any image — from the homepage, from a collection, or from a direct link — and explore what's connected to it. A mosaic of related images appears, each one a doorway to more. Switch between five dimensions of similarity to see completely different perspectives on the same starting point.
Beyond keyword search
AI that understands images the way you do
At the heart of Endlss is CLIP — a neural network that generates 512-dimensional embeddings for every image. These embeddings capture what an image looks like, feels like, and means — not just its surface-level tags. Two images can share no keywords in common and still be deeply connected through visual style, mood, or composition.
Combined with colour palette analysis, community-curated tags, and prompt-based similarity for AI art, every image becomes a hub of multi-dimensional discovery.
Visual similarity
Powered by CLIP embeddings — an AI model that understands images the way a human eye does. Find images that share composition, style, subject matter, and mood.
Colour palette
Every image is analysed for its dominant colours. Explore by palette to find images that share your favourite tones, gradients, and chromatic moods.
Tags
AI-generated and community-curated tags let you explore by subject, style, and theme. Vote on tags to improve discovery for everyone.
Prompt similarity
For AI-generated images, explore by the text prompts that created them. Find images born from similar creative ideas and directions.
Name and title
Discover connections through image titles and filenames. Sometimes the creator's naming reveals thematic links that visual analysis alone wouldn't catch.
Combined discovery
Switch between dimensions on any image to see completely different sets of related content. One image, five perspectives — each revealing new connections.
Frequently asked questions
- How does the image explorer work?
- Select any image on Endlss and you'll see a mosaic of similar images — connected by visual similarity, colour, tags, prompts, or names. Select any of those images to go deeper. Every path through Endlss is unique.
- What technology powers the image exploration?
- Endlss uses CLIP (Contrastive Language-Image Pretraining) to generate 512-dimensional embeddings for each image. Images with similar visual content are close in this embedding space, enabling nearest-neighbour similarity search that understands style, composition, and subject matter.
- Can I explore images by colour?
- Yes. Every image on Endlss is analysed for its dominant colour palette. You can explore by colour to find images that share specific tones and moods, or use the dedicated Explore by Colour page to start from a specific hue.
- Is the image explorer free?
- Yes. Browsing and exploring images on Endlss is completely free. You can discover, follow threads of similarity, and build collections without any cost.
Interested in video? Endlss also supports video clip exploration. Or discover how creatives use Endlss for artistic inspiration.