Launched this week

LocalXiv

Launched this week

Research papers, explained visually.

4 followers

Research papers, explained visually.

4 followers

Visit website

Productivity

•

Note and writing apps

•

Product demo

LocalXiv turns arXiv and alphaXiv links into a paper library on your Mac. Read with adjustable fonts and layouts, explore visual overviews and cited explanations, and export papers or explanations for Kindle. Free and open source. Reading and export need no AI account; optional AI uses your own provider and API key.

Free

Launch tags:Productivity•Open Source•Developer Tools

Launch Team / Built With

OpenAI

GPT-6 Astra Challenge

How did Astra change the scope or ambition of what you built?

Maker

LocalXiv started as a simple Chrome extension: take an arXiv paper and send it to a Kindle. As I worked on it, I wanted to improve the reading experience itself, not just how the paper reached the device. Astra helped me plan the project, design the interface, and brainstorm and refine the workflows. That support made it feasible to expand the extension into a local Mac app with a paper library, adjustable reading layouts, visual overviews, and longer explanations, while keeping Kindle export. Without Astra, I would likely have stopped at the original extension. Having help with both planning and implementation let me take on a broader product. The goal changed from “send this paper to my Kindle” to “help me read and understand this paper in the format that works for me.”

Report

2d ago

Maker

📌

Hi, I’m Akhil, and I worked on LocalXiv.
I built it to make research papers easier to understand and more comfortable to read. You can start with a visual overview, move to a longer explanation with source references, and then read the paper itself in a customizable reader.

It also supports exports for reading outside the app, including on Kindle.
This is an early macOS release. Your library is stored locally; optional AI features use your selected provider and API key.

This is an early preview. Conversion can be imperfect, and AI explanations can contain mistakes. The Mac build is not notarized yet; first-launch instructions are in the repository. I’m particularly interested in feedback on whether the explanations preserve the paper’s actual contribution and limitations, and where the import-to-reading experience becomes confusing.

Report

3d ago