Introducing quickarXiv Papers are often written in convoluted language that is hard to understand. We're fixing that. Swap arxiv → quickarxiv on any paper URL to get an instant blog with figures, insights, and explanations. Now extracted with DeepSeek OCR 🚀

Example:

this is so cool. LFG

Tried doing yesterday. Glad this just rolled out 😆

ok so we're doing papiers again but with less features "papers are written in convoluted language" is a take that reveals you're not the target audience papers are dense because: they need to be precise (ambiguity = unreproducible) they assume domain knowledge (that's what citations are for) they follow conventions that let experts skim efficiently page limits force compression "convoluted" = "I don't have the background to read this efficiently" which is fine! but the solution isn't dumbing it down, it's building up your knowledge what quickarxiv actually does: OCR the PDF (DeepSeek, fine) GPT-wrapper summarization extract figures reformat as blog this already exists: papiers (we just talked about this) explainpaper arxiv-sanity literally just pasting into Claude the "blog format" is actively worse for some content: math derivations need the actual math methodology sections need precision results need the actual numbers and confidence intervals "insights and explanations" means "LLM's interpretation of what matters" the failure mode: you read the quickarxiv blog, think you understand the paper, cite it in your own work, get the details wrong because the summary elided something important. this already happens constantly also "DeepSeek OCR" as a feature is funny because arxiv has had machine-readable source files forever. most papers have LaTeX source. you don't need OCR, you need to parse the .tex who this is actually for: people who want to seem like they read papers without reading papers. VCs, journalists, twitter threads. not researchers tldr: it's an LLM summary with figures extracted. useful for skimming, dangerous for understanding. the "papers are convoluted" framing tells on itself — the density is a feature not a bug

Cool!

Is it possible to tweet the prompts for blog generation (say, adding explanations to some parts of the paper), and downloading the blog as markdown?

Highly recommend providing side-by-side translations of the papers themselves, with English and translated languages displayed in parallel! An absolute must-have

@BrianRoemmele

~ thanks👍🏾

love this for a dummy like me

Are you open sourcing this?

Now do quickTwitter. So much research being described with influencer terms and put into N post threads.

what if you treat comprehension in itself as a learning experience rather than trying to circumvent everything

huge for digesting info in a more understandable manner for everyone, huge W thanks for pushing

Cool!

Looking forward to the final product! I've always felt like I'm missing a way to recommend papers and get quick summaries.

Interesting idea! I'll try it out.

I'm not sure if its the cloudflare outage affecting Google SSO but i'm unable to use this service

LOVE this

🙏👍

demo looks cool! is this open source?

Sound like a pretty decent project

this just what I need. does it support discussion with multiple papers instead just one or full comparison in the library

@torchborg quickarXiv for dense papers is huge! What’s been the most convoluted paper you’ve tackled recently? 🤔

Found another app going after this concept a couple days ago and really liked how easy it made digesting articles. Going to give this a go next!

noob version, i like it.😍

Can you make it work for biorxiv?

cook

Super cool, would love to collab on a new 100% AI research faculty

The UI design is so good. Also the browser plugin is pretty useful. Nice works!

where can I see papers I liked? seems bookmarks are only shown

This is so cool

neat

@askalphaxiv Can I somehow integrate this feature into my Zotero collection?

Doesn't work on mobile?