I've built Fractal Reader, a tool that creates zoomable, multi-level AI summaries for articles and academic papers.

https://www.fractal-reader.com

Problem: Existing summarization tools typically offer a single summary, which is often too vague or too detailed.

Solution: Fractal Reader generates multiple zoomable summaries—from a quick one-sentence overview to detailed paragraph-level summaries—allowing you to seamlessly zoom in or out on content based on your needs. Quickly grasp the big picture, then selectively explore deeper insights without re-reading entire sections.

The demos probably illustrate it best:

Summary of "Attention Is All You Need" Paper

https://www.fractal-reader.com/view/d9bccd98-8580-4c0a-a366-99b0b73a865e

Summary of "Do things that don't scale" Essay by Paul Graham

https://www.fractal-reader.com/view/184e28f3-0540-4368-a654-3dd0d875ba90

Try it out with your own articles or papers here:

https://fractal-reader.com/

---

Under the hood: Fractal Reader leverages Gemini-based summarization in a custom pipeline that intelligently segments text and constructs hierarchical summaries. It's built to smoothly handle both general articles and dense academic material.

I'd love your feedback:

- Does this solve a real pain point for you? - Can you think of other scenarios or improvements that would make this tool even more valuable?