Help·What is Slurp?
What is Slurp?
Automated discovery
How Endlss finds content at scale
Slurp is our automated content harvester. It connects to public sources across the web — subreddits, image communities, creative platforms — and collects images and videos that might belong on Endlss.
It runs continuously, pulling in thousands of pieces of media every day. That scale is what allows Endlss to offer such a broad and varied library of visual content, but it also means Slurp is, by nature, indiscriminate. It casts a wide net.
How it works
The pipeline from source to Endlss
1. Harvesting
Slurp connects to configured sources using their public APIs. It fetches recent posts, extracts image and video URLs, and collects metadata like titles, authors, tags, and prompts where available. Slurp is configured to collect more media that matches media that is visited and liked on Endlss, so if we have a lot of views and likes on a picture of a tomato plant, Slurp tries to get more pictures of tomatoes and of plants and tomato plants and everything inbetween. There is a balance to be struck here, because we don't want too many tomatoes, but we also want to give Endlss visitors more of what they came to see.
2. Downloading and processing
Each piece of media is downloaded, thumbnailed, and processed. We extract dominant colours, generate CLIP embeddings for similarity search, and catalogue the content automatically.
3. Moderation
Every item Slurp collects goes through moderation before it goes live on Endlss. This includes automated AI filters for quality and content safety, as well as manual human review. Content that doesn't meet our content guidelines is rejected.
4. Publishing
Content that passes moderation is published to Endlss, tagged, themed, and made discoverable through our similarity engine. The original source URL and where available, the author, is always preserved.
The reality of scale
Why moderation matters
Slurp collects thousands of images and videos. At that volume, not everything it picks up is appropriate, high quality, or correctly attributed. Some content may have been posted to public platforms without the original creator's permission. Some may be low-effort or AI-generated slop. Some may simply not be interesting.
We moderate everything that comes through Slurp before it reaches you. Our moderation pipeline combines automated AI filtering with human review to catch content that doesn't belong. But no system is perfect — if something slips through, we want to know about it.
Slurp respects each platform's public API, rate limits, and robots.txt rules. It only accesses publicly available content.
Your content and Slurp
Removal, attribution, and exclusion
If your work was collected by Slurp and you want it removed, corrected, or you'd like to be excluded from future collection, see our detailed Slurp policy or read When things go wrong.
You can also email takedown@endlss.co at any time and we will act promptly.