Open Source - Page 5 of 21
Open-source AI projects, community innovations, collaborative development, and freely accessible AI tools and frameworks.
406 articles
View complete article list
Qwen3.5 9B MTP Tops Local Coding Models for Scripts, Debugging and Assistants
Everyone's racing to stuff an AI model onto your laptop. Most of them are useless at the actual job of programming.
Mistral OCR 4 Delivers Citation‑Ready Structured Output for RAG and Search
Data is the lifeblood of modern AI, but raw document text is a hemorrhage of noise. Mistral OCR 4 staunches that flow by delivering structured output so clean it’s practically citation-ready, no post-processing, no guesswork.
Krea 2 Raw/Turbo generate AI images in 2 s; Nano Banana Pro 17.7 s proprietary
Two seconds. That’s all it takes for Krea 2 Raw and Turbo to generate an enterprise-grade image.
GLM-5.2 API guide emphasizes tool‑based lookups, not guesswork
Guesswork is a luxury no serious analyst can afford. Numbers demand precision, not approximation. The GLM-5.2 API guide cuts through the fog: tool-based lookups replace guesswork. No more hallucinating populations or fumbling arithmetic.
AI research increasingly relies on recursive loops, a staple of CS basics
Every comp-sci freshman learns recursion: a function calls itself, cracks a problem, and stops. That's the textbook version. Step inside an AI lab today, and that neat, predictable structure has shattered. The loops have gained a will of their own.
Amazon MGM shelves Luca Guadagnino’s Sam Altman film starring Andrew Garfield
The week Sam Altman was fired from OpenAI and then reinstated was a masterclass in chaos, a corporate thriller played out in real time across five dizzying days.
Anthropic shuts down Fable 5 and Mythos 5 models amid White House dispute
The shutdown of Fable 5 and Mythos 5 didn’t just kill two advanced AI models, it detonated a political bomb. Already hemmed in by safeguards limiting their use in high-risk areas, the models were yanked completely after a White House dispute.
Mistral AI seeks EUR 3 bn, valued at EUR 11.7 bn; ASML holds 11% stake
Paris-based Mistral AI is now chasing a staggering three billion euros in fresh capital. Look past the startup facade.
Dario Amodei has one direct report; sister Daniela runs Anthropic's exec team
Dario Amodei has one employee. The CEO of a $30 billion company directly manages exactly one person. His co-founder, Daniela Amodei, runs everything else. This isn’t an accident. It’s a philosophy made concrete.
GPU utilization masks storage and I/O bottlenecks slowing modern AI
The GPU’s own metrics are lying to you. They flash high numbers, near-100% occupancy, impressive compute throughput, but underneath, the system is rotting.
Matmul Enables Dropless MoE Training; Grouped‑GEMM Kernel Drives Speed
Dropless Mixture-of-Experts training has long been haunted by a simple truth: you cannot skip the tokens, but you can pay for them in compute.
LangChain Emergency Helpline Uses AssemblyAI WebSocket for Live STT
The caller’s voice arrives in raw audio chunks, each one a fragment of a crisis. Milliseconds matter.
Meta launches Hatch AI agent, its first paid product, priced up to USD 200/month
Two hundred dollars a month. That’s the price of a decent car payment, or a generous grocery run. And it's exactly what Meta is asking for Hatch, its first paid AI. Forget the chatbots. You describe a tool in plain English, and this thing builds it.
Open‑source voice model listens continuously, decides to speak every 0.4 seconds
Most voice assistants are terrible listeners. They wait for a magic word, ignore the noise in between, and treat a conversation like a series of disconnected commands.
NSA uses Anthropic's Mythos AI model for offensive cyber ops against China, Iran
The National Security Agency is hacking foreign governments with a corporate AI. It’s adapting Anthropic’s Mythos model for cyber attacks on China and Iran.
TensorFlow Emotion Dataset with 54,263 Texts Shows Class Imbalance
54,263 labeled texts. That’s the raw size of the TensorFlow Emotion Dataset, a resource released under Apache 2.0 for anyone building emotion classifiers. Look closer, and a stark pattern emerges: neutral examples dominate. The rest?
xAI launches Grok Imagine 1.5, adding 720p text‑prompted image‑to‑video
OpenAI’s Sora is stuck in the garage. xAI just rolled out a new car. The company launched Grok Imagine 1.5, a model that turns a single picture into a short, 720p video. You tell it what to do with text prompts: pan left, speed up, make it moody.
Ideogram 4.0 releases open-weight 2K model, GitHub weights, tops DesignArena
You can download the model right now. You can also build on its code. But if you want to make money from it, you’ll have to pay. That’s the central tension of Ideogram 4.0, an open-weight image generator that just became the strongest of its kind.
Nous Research releases Hermes Desktop, cross‑platform GUI for Hermes Agent v0.15.2
Most AI agents live in the terminal, a place of cryptic commands and invisible thought. Nous Research just gave theirs a face. They released Hermes Desktop, a native graphical interface for their Hermes Agent v0.15.2. It works on Windows and Mac.
AI frees time of laid‑off workers as studios seek actors to replace theater kids
The promise was always free time. Your life, returned. Then the layoffs at animation and effects studios hit. Now, as those same companies audition actors before an AI can do it cheaper, the irony is a tangible thing.
Featured Resources & Reviews
No Code MBA Course Review
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
AI Tools & Resources
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
Weekly AI Digest
Get the week's most important AI news delivered to your inbox every week.