Open Source LLM

DeepSeek pitches new route to scale AI, but researchers call for more testing

DeepSeek’s paper is fuelling speculation that its next models could incorporate new architecture. Photo: Reuters

DeepSeek’s proposed “mHC” architecture could transform the training of large language models (LLMs) – the technology behind artificial intelligence chatbots – as developers look for ways to scale models without simply adding more computing power.

However, experts cautioned that while the approach could prove far-reaching, it might still prove difficult to put into practice.

In a technical paper released last week, co-authored by DeepSeek founder and CEO Liang Wenfeng, the

Open Source LLM

Tech war: China takes confident strides to develop more AI innovation in 2026

Illustration: Henry Wong

That release was a fitting reminder to the world, especially during the peak of the Christmas holiday season, about Chinese AI companies’ sharpened focus on innovation to stay ahead in this fast-developing industry.

It was around this time last year when DeepSeek started to get widely noticed days after releasing its namesake large language model (LLM), DeepSeek-V3 . Weeks later on January 20, reasoning model DeepSeek-R1 was released

Open Source LLM

Chinese start-ups Zhipu and MiniMax release latest AI models ahead of Hong Kong listing

Zhipu AI, known internationally as Z.ai, says its new GLM-4.7 model features improved coding abilities. Photo: Shutterstock

The near-simultaneous release by two of China’s “AI tigers” underscored efforts to drum up interest for their respective initial public offerings in Hong Kong , after both companies passed crucial regulatory hearings over the weekend.
Beijing -based Zhipu AI, known internationally as Z.ai, on Monday evening claimed that its new GLM-4.7 model features improved coding abilities, narrowing the gap with larger US firms Google DeepMind and Anthropic in

Open Source LLM

OpenAI rolls out a fun ‘Your Year with ChatGPT’ feature in select countries

OpenAI has joined the Spotify Wrapped -style personalized year-end recap trend with “Your Year with ChatGPT,” which is rolling out to users in the U.S. and select additional countries. Here’s how to see yours.

In a post on X earlier tonight, OpenAI confirmed that “Your Year with ChatGPT” is rolling out now to users in the US, UK, Canada, New Zealand, and Australia who are running the latest version of the ChatGPT app.


Open Source LLM

Lua 5.5 Released With Declarations For Global Variables, Garbage Collection Improvements

Five years after releasing Lua 5.4, Lua 5.5 debuted today as the newest version of this lightweight and embeddable scripting language.

Lua 5.5 brings support for declarations of global variables, more compact arrays, and other refined language features. Large arrays in Lua are as using around 60% less memory with the Lua 5.5 build. Lua 5.5 also continues work on garbage collection "GC" improvements with a new generational mode and major garbage collections being done in an incremental manner.
Continue reading...

Open Source LLM

Indie App Spotlight: ‘AnywAIr’ lets you play with local AI models on your iPhone

Welcome to Indie App Spotlight . This is a weekly 9to5Mac series where we showcase the latest apps in the indie app world. If you’re a developer and would like your app featured, get in contact .

, which is a play on the word “anywhere”, is a nifty little iOS app that lets you play with AI models – regardless of if you have an internet connection. It offers custom themes, a plethora of tools and games, and all of the local AI models you could

Continue reading...

Open Source LLM

How NotebookLM made self-hosting an LLM easier than I ever expected

A low-friction path to setting up my local LLM

Open Source LLM|3D print

Prusa Research introduces the Open Community License to protect open source 3D Printing hardware — new rules aimed at addressing industry abuses

(Image credit: Tom's Hardware)
Follow us

Josef Prusa has long been a champion of Open Source, but defending its use has been an uphill battle. Last July, he surprised no one at all when he announced that “ open hardware desktop 3D printing is dead .” He pointed to his company’s 2016 MMU1 multiplexer as an example of an open source design

Continue reading...

Open Source LLM

OpenAI發布新模型 付費ChatGPT用戶即日起可使用

在發布GPT-5.2系列模型一周後,OpenAI於18日推出Codex模型GPT-5.2-Codex,號稱迄今為止最先進的智能體編程模型,專為復雜現實場景的軟件工程而設計。即日起,付費ChatGPT用戶可在所有Codex界面使用GPT-5.2-Codex,其正努力在


Open Source LLM

「Gemini 3 Flash」提供開始。2.5 Proより速く賢い新モデル

Googleは、高速かつ低コストな新LLM(大規模言語モデル)「Gemini 3 Flash」を発表し、開発者向けに提供開始した。
Menu