Top Tweets for #Building
Conquering the world on horseback is easy. It is dismounting and governing that is hard.
— Kublai Khan,
Winning an empire takes strength. Building one that lasts takes vision. That is the mark of LEGACY.
$LEGACY
2XnP5fdNbeBbBUX1sh4gKTJCM7zD1SoKX2QhEHudpump

Full-stack engineer turned GTM engineer, starting X from 0 followers.
Next 100 days: building the system that has to bring in $1M in closed deals for Thinkiyo, in public.
Engineers: would you build it (code, n8n, Supabase) or buy tools like Clay and glue them together?
🌎👀 I’m building a community, not just an audience! Join THE WORLD WATCHERS and be part of the conversations, debates, laughs, and everything happening in the world.
Don’t just watch the world. Be part of it! 💯
#TheWorldWatchers #TheWorldWeLiveInPodcast
This is next-level AI building. The possibilities are endless!
So I built a 3D Game 🎮
Neon Cyber-Drift: 100% procedural 3D synthwave racer one-shotted with OpenClaw + Kimi K3.1 Pro @cerebras.
Pure Three.js & Web Audio, 0 assets.
Play: neon-cyber-drift.vercel.app
Code: github.com/JimmyOgb/Kimi-k3.…
@Kimi_Moonshot #KimiOnCerebras
ICM 1/35th scale U.S. Army Standard type B "Liberty Truck". This is a really nice kit and I enjoyed building it more than I expected.
#JFFWWI

This is next-level AI building. The possibilities are endless! 🔥
So I built a 3D Game 🎮
Neon Cyber-Drift: 100% procedural 3D synthwave racer one-shotted with OpenClaw + Kimi K3.1 Pro @cerebras.
Pure Three.js & Web Audio, 0 assets.
Play: neon-cyber-drift.vercel.app
Code: github.com/JimmyOgb/Kimi-k3.…
@Kimi_Moonshot #KimiOnCerebras
Hi all! I'm Jeremy, a web developer in Shrewsbury, MA, building browser games with three.js.
My latest: JAT Arcade Archives, a 3D museum of 41 playable machines from 1947 to 1989.
Would love feedback on how it runs for you. Link in the reply

Keep building! I'd be happy to explore a collaboration. 🔥
Idgaf about the sharking content fr cause Cinco gon learn the hard way that the people watching that arent Jay cinco fans fr but its just annoying to see the content regress. Spent the yr building a solo brand successfully just to go back to the same old streams & broken promises
Hi Iris! Which are you focused on this week, Adant or Atlas? I'm building TeaQL to give coding agents typed query/mutation APIs generated from business models.
Respect the consistency. Keep building and showing up
Always refreshing to see developers building something just for the joy of it
sunday afternoon, still shipping.
let's connect if you're:
- making films or ads with ai video
- coding with claude code or cursor
- into maps, history or data viz
- building for kids
i'm iris, building adant (ai video) and atlas, a history globe. reply with what you're making
There is now ceiling!
We keep building.

Building a local AI agent that fixes itself. Today, in a live run, ours caught itself going in circles.
Her watch named the stall, rewound the work to the spend that caused it, marked that road refused, and she resumed differently.
Nobody touched the keyboard. That's the job.
If you’re looking for a different kind of crypto community to discover today, take a look at what $WISHMAS is building. 🎄
🛠️🧭 I went down the OCR + PDF extraction rabbit hole.
If you are building RAG systems or AI Agents around real documents, you also have to deal with:
tables, formulas, reading order, scanned pages, multi-column layouts, charts, headers and footnotes.
So I went through the open-source ecosystem and pulled together the tools I would actually look at.
Here are 20 OCR + PDF extractors worth knowing in 2026:
》𝗡𝗮𝘁𝗶𝘃𝗲 𝗣𝗗𝗙 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻
✸ PyMuPDF4LLM → PDF to Markdown/JSON
✸ pdfplumber → text, coordinates, tables
✸ pypdf → lightweight PDF parsing
→ Best starting point when the PDF already contains clean digital text.
》𝗢𝗖𝗥 𝗘𝗻𝗴𝗶𝗻𝗲𝘀
✸ Tesseract → classic OCR baseline
✸ PaddleOCR → multilingual OCR + document AI
✸ Surya → OCR + layout + reading order
✸ EasyOCR → simple Python OCR
✸ docTR → text detection + recognition
》𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁 𝗣𝗮𝗿𝘀𝗲𝗿𝘀
✸ Docling → PDFs, Office, tables, layout
✸ MinerU → formulas, tables, reading order
✸ Marker → PDF/image to Markdown, JSON, HTML
✸ Unstructured → extraction + chunking
》𝗩𝗟𝗠-𝗕𝗮𝘀𝗲𝗱 𝗣𝗮𝗿𝘀𝗶𝗻𝗴
✸ olmOCR → VLM-based PDF extraction
✸ OCRFlux → complex layouts + cross-page tables
✸ MonkeyOCR → multimodal document parsing
✸ dots.mocr → multilingual structured documents
✸ Zerox → vision models to Markdown
→ This is the category I find especially interesting.
The model is not only recognizing characters anymore.
It is also interpreting the document structure.
》𝗦𝗽𝗲𝗰𝗶𝗮𝗹𝗶𝘀𝘁 𝗧𝗼𝗼𝗹𝘀
✸ GROBID → scientific papers + metadata
✸ Camelot → PDF tables
✸ Nougat → equations + academic PDFs
And this is the bigger shift:
PDF
→ Native text
→ Layout
→ OCR / VLM
→ Tables + formulas
→ Reading order
→ Markdown / JSON
→ Verification
→ RAG / Agent context
𝗣𝗹𝗲𝗮𝘀𝗲 𝗥𝗲𝗺𝗲𝗺𝗯𝗲𝗿:
The best OCR model is not necessarily the best document pipeline.
For Production AI Agents, the goal is not only to extract the words.
It is to preserve enough structure and meaning for the agent to reason over the document correctly.
I may benchmark Docling vs MinerU vs Marker vs olmOCR vs PaddleOCR vs OCRFlux next.

Mike green we have hope in you next year keep building
What if you could start with a tiny island, turn it into a full pirate harbor, build a fleet of ships, and send them on real-time treasure voyages while you sleep? 🏴☠️
That’s @PlunderPort — a pirate idle tycoon + NFT adventure building toward Zero (LayerZero ecosystem).
Institutions don’t stay offchain because of fees.
They stay offchain because their entire book is live-streamed to competitors.
AI agents have the same problem
zkTLS is the missing primitive: private where it matters, provable where it counts.
@primus_labs is building it

Currently workflows for some things I’m building around Decision models , VLMs and whatever works .
Basically want the AIs from the movies we all grew up with :)
Keep building
Both approaches serve different purposes, giving developers more flexibility when building onchain applications. Not every transaction follows the same schedule, so execution should not have to follow just one approach.
Yes Fringe strong body strong stack keep building
Context is everything when building with AI
Building something simple that people actually need is always a good move
The reach Is there. Now it's about improving content quality and building stronger connections
Most Popular Users

Elon Musk 
@elonmusk
240.3M followers

Barack Obama 
@barackobama
118.6M followers

Cristiano Ronaldo 
@cristiano
114.4M followers

Donald J. Trump 
@realdonaldtrump
111.8M followers

Narendra Modi 
@narendramodi
106.9M followers

Rihanna 
@rihanna
98.7M followers

NASA 
@nasa
92.2M followers

Justin Bieber 
@justinbieber
91.7M followers

KATY PERRY 
@katyperry
89.9M followers

Taylor Swift 
@taylorswift13
83.9M followers

Lady Gaga 
@ladygaga
75.4M followers

Virat Kohli 
@imvkohli
73.3M followers

Kim Kardashian 
@kimkardashian
70.8M followers

YouTube 
@youtube
68.6M followers

Neymar Jr 
@neymarjr
66.3M followers

Bill Gates 
@billgates
65.2M followers

Selena Gomez 
@selenagomez
63M followers

The Ellen Show
@theellenshow
62.1M followers

CNN 
@cnn
61.7M followers

X 
@x
60.6M followers


























