Unclecode (Hossein) (@unclecode) / X

Unclecode (Hossein)

2,611 posts

Unclecode (Hossein)

@unclecode

Author of Crawl4AI (#1 GitHub Trending). Founder of Kidocode, SE Asia's largest tech & biz school. Leading AI research on synthetic data, AI consultant.

Singapore

unclecode.com

Born March 22

Joined September 2009

Unclecode (Hossein)
@unclecode
May 9, 2024
Crawl the web in an LLM-friendly style! Introducing Crawl4AI 🤖🕷️which is a web data crawler that extracts semantically labeled chunks into JSON, along with clean HTML and markdown for RAG, fine-tuning, and AI chatbots. This open-source tool offers efficient crawling and
00:00
111K
Unclecode (Hossein)
@unclecode
May 1, 2025
Replying to @sama
That special hard drive is called “hugging face” and that can be done today no need future.
17K
Unclecode (Hossein)
@unclecode
May 17, 2024
Crawl the web in an LLM style! Exciting to share the Crawl4AI v0.2.0 with significant improvements, for everyone care for a good crawler. - 🚀 10x faster! - 📜 Execute custom JavaScript - 🤝 Colab friendly! - 📚 Chunking strategies - 🧠 Extraction strategies: cosine
00:00
31K
Unclecode (Hossein)
@unclecode
Jun 24, 2024
🎉 Craw4AI is almost at 1,000 stars on GitHub! ⭐️ v0.2.7, Crawl the web in LLM-friendly Style. Lots of changes have been made: hooks, events, structured data extraction, vision, and more. The biggest improvement is speed. In this video, I explain these changes and my future
00:00
50K
Unclecode (Hossein)
@unclecode
Jul 20, 2025
Amazing article by @rasbt comparing 6 LLMs Architecture, this Kimi 2 @Kimi_Moonshot is phenomenal, 1 TRILLION → 54B active! MoE is wild - more knowledge, same compute. I made a Claude artifact, you can play with it to compare models, plus a glossary of terms used in this
15K
Unclecode (Hossein)
@unclecode
Mar 3, 2024
Function call for Groq 😎 Hey @GroqInc I've developed a proxy server enabling function calls for Groq cloud, couldn't wait for you guys :D, and it has OpenAI interface compatibility. It works seamlessly with libraries like LlamaIndex, LangChain, and my favorite, PhiData. Check
00:00
41K
Unclecode (Hossein)
@unclecode
Apr 26, 2024
LLama3 + Function Calling I've successfully fine-tuned Llama3 to natively support function-calling, and I did the same for TinyLlama, which also worked surprisingly well! I'm motivated to try this with a model under 1B parameters next! That's my upcoming challenge. I plan to
00:00
26K
Unclecode (Hossein)
@unclecode
Oct 29, 2024
🚀 Crawl4AI v0.3.72 is here! 🚀 What’s New? ✨ Fit Markdown: Extracts only the main content from articles and blogs 🛡️ Magic Mode: Comprehensive anti-bot detection bypass 🌐 Multi-browser support: Switch between Chromium, Firefox, WebKit 🔍 Knowledge Graph Extraction:
00:00
15K
Unclecode (Hossein)
@unclecode
Aug 6, 2024
🚀 GroqChat: An app that deserves to become the new chat.groq.com! 😉 I whipped up GroqChat - a lightning-fast, browser-based chat interface powered by @GroqInc LPU. It's privacy-first, @vercel hosted, and Groq-exclusive. Features: 🏠 100% local, runs in your browser
00:00
23K
Unclecode (Hossein)
@unclecode
Aug 2, 2024
🚀 𝐂𝐫𝐚𝐰𝐥𝟒𝐀𝐈 𝐯𝟎.𝟐.𝟕𝟔 𝐢𝐬 𝐡𝐞𝐫𝐞, 𝐚𝐧𝐝 𝐢𝐭'𝐬 𝐨𝐮𝐫 𝐜𝐨𝐨𝐥𝐞𝐬𝐭 𝐫𝐞𝐥𝐞𝐚𝐬𝐞 𝐲𝐞𝐭! 🎉 ✨ New features: 🐳 Official Docker Hub image l 🖼️ Image description generation 🌐 Improved cross-platform compatibility ⚡ Major performance boost And more …. Huge
00:00
19K
Unclecode (Hossein)
@unclecode
Jul 30, 2025
Crawl4ai crossed 50K start 🌟 (~5k forks). Thanks to all who shared their love and support. A project I started to build a tool for my company turned into a lovable open-source project at this scale. It's so rewarding to be part of an authentic change. I like to believe Crawl4ai
18K
Unclecode (Hossein)
@unclecode
Feb 21, 2025
Replying to @deepseek_ai
You are a tiny team but enough to break the giant AI dominancy. As the author of Crawl4ai and anotyer tiny team, I really hope there is a way I can contribute in your causes which aligned deeply with my core values.
16K
Unclecode (Hossein)
@unclecode
Jul 21, 2024
Replying to @fchollet
That's a scary misunderstanding! A crucial point: Occam's Razor, also called 'Principle of Parsimony', aims for the most parsimonious explanation, not just the simplest. It seeks the explanation with fewest assumptions that still accounts for all phenomena. It's not about
4.7K
Unclecode (Hossein)
@unclecode
Aug 29, 2024
Replying to @rohanpaul_ai
You have to make an API, and then it’s a paid service, why call it open source? Have you tried Crawl4ai? It’s extremely fast because it's smart enough not to overuse LLMs! Here, you pay for the LLM and the API!! Crawl4ai also provides structured JSON, which you can use with any
6.3K