The AI industry's infrastructure layer is being acquired at breakneck speed. Nvidia has agreed to buy Hugging Face for $12.9 billion, Stripe finalized a $7.5 billion deal for OpenRouter, and Anthropic's annualized revenue hit $65 billion - up from $9 billion just nine months ago. These three deals alone total over $85 billion and signal that the era of independent AI infrastructure companies is ending.
AI & ML News
Nvidia Nears $13 Billion Deal for Hugging Face
Nvidia has agreed to acquire Hugging Face, the world's most popular open-source AI model hub, for $12.9 billion, according to The Information. The deal would value the company at nearly triple its $4.5 billion Series D valuation from 2023. Hugging Face was generating approximately $150 million in annual revenue as of late August, up from $100 million two months earlier, and was approaching profitability. The acquisition gives Nvidia a dominant position in the open-source AI ecosystem at a critical moment - the company has been pushing to protect open models amid Washington debates about restricting open-weight AI systems. Hugging Face CEO Clem Delangue had publicly aligned with Nvidia's open-source advocacy, appearing on CBS's "Face the Nation" and signing a letter with 24 companies urging the U.S. government to support open models. The deal also gives Nvidia a path back into cloud computing, as Hugging Face already helps developers run models using rented compute power.
Stripe Closes $7.5 Billion Acquisition of OpenRouter
Stripe has finalized its acquisition of OpenRouter, an AI model routing startup, for $7.5 billion, the New York Times reported. Of that sum, $1.5 billion goes to founders and $6 billion to investors. The deal represents a stunning 5.4x markup over OpenRouter's $1.3 billion valuation from its Series B round in May 2026. OpenRouter serves 8 million global users and provides access to more than 400 AI models, routing requests to the optimal provider based on cost and performance. CEO Alex Atallah had described the company as "the Stripe for AI" - a comparison that proved prophetic when Stripe came calling. The acquisition gives Stripe a foothold in AI infrastructure as businesses face mounting pressure to manage AI costs across multiple providers.
Anthropic Revenue Surges to $65 Billion Annualized Run Rate
Anthropic's annualized revenue surpassed $65 billion at the end of July, up from $47 billion in May and just $9 billion at the end of 2025, Bloomberg reported. Investors expect the company to finish 2026 with revenue between $100 billion and $120 billion, according to the Financial Times. Both Anthropic and OpenAI have filed confidential IPO paperwork, with Anthropic expected to hit public markets first - possibly this fall - seeking a valuation of $2 trillion or more, which would make it the largest market debut on record. OpenAI, by comparison, has doubled its revenue to $40 billion. Anthropic was last valued at $965 billion in late May when it raised $65 billion. The company's growth has been driven largely by its AI coding capabilities through Claude Code and Cowork platforms.
Apple Caught Off Guard by AI Demand for Mac Mini and Mac Studio
Apple announced new Mac Mini and Mac Studio models launching September 22, but the bigger story is that the company was unprepared for demand. CEO Tim Cook said on the Q2 earnings call that AI developers and OpenClaw users have been buying up existing models faster than Apple predicted. "Both of these are amazing platforms for AI and agentic tools, and the customer recognition of that is happening faster than what we had predicted," Cook said. The new Mac Mini starts at $899 with M6 and M5 Pro chips, while the Mac Studio gets M5 Max and M5 Ultra options with up to 512GB of memory. Apple warned that supply constraints from chip manufacturing and memory shortages could persist for several months. IDC expects PC shipments to decline 11.3% in 2026 due to the global memory shortage.
Groq Raises $350 Million to Complete Pivot from AI Chips to Neocloud
Groq has raised $350 million at a $3.5 billion valuation as it completes its transformation from an AI chipmaker to a neocloud provider operating Nvidia systems. The funding round was led by Disruptive with planned participation from Nvidia. Groq now operates 13 data centers across North America, Europe, the Middle East, and Asia Pacific, serving more than 6 million developers. The valuation is down from $6.9 billion last September, but the company frames it as a reset for its "post-Nvidia-licensing-deal version." In June, Nvidia hired Groq's founder Jonathan Ross and top talent as part of a $20 billion licensing deal, forcing the pivot. Groq intends to scale from 54 megawatts to more than 200 megawatts in 2027.
Big GitHub Releases
RavynOS: Pre-Alpha Open-Source OS Based on Darwin and FreeBSD
RavynOS has reached pre-alpha status as an open-source operating system built on Darwin, FreeBSD, and Apple open-source components. The project gained 175 points on Hacker News with 106 comments, signaling strong community interest in a macOS-compatible open-source desktop. The project aims to provide a macOS-like experience using open-source foundations, allowing users to run macOS software on non-Apple hardware. It joins DarlingHQ, a Darwin/macOS emulation layer for Linux, which also trended on GitHub and HN the same day with 134 points and 42 comments. Together, these projects represent a growing movement to bring macOS software compatibility to open platforms.
Chrome DevTools MCP Reaches 50,000 Stars for AI Agent Browser Automation
Chrome DevTools MCP has hit 50,054 GitHub stars, gaining 67 stars today, making it one of the most popular tools for giving AI coding agents browser automation capabilities. The project provides a Model Context Protocol server that lets agents interact with Chrome's DevTools protocol for automated testing, web scraping, and browser-based development workflows. Built by the Chrome DevTools team, it bridges the gap between AI coding assistants and real browser environments. The tool supports automated testing, performance analysis, and debugging through natural language interfaces.
Dwarf Fortress Announces Largest Magic Update in History
Dwarf Fortress is getting what creators Tarn and Zach Adams describe as "the mother of all magic updates," extending to the fundamental cosmological makeup of the universe. The update, which received 316 points and 112 comments on Hacker News, will add magic systems that interact with the game's legendary procedural world generation. Dwarf Fortress has been in development for over 20 years and is considered one of the most complex simulation games ever created. The magic update represents the most significant content addition since the game's Steam release.
Under-the-Radar Finds
ponytail: AI Agent That Thinks Like the Laziest Senior Dev (115K Stars, +1,396 Today)
ponytail has exploded to 115,872 GitHub stars, gaining an extraordinary 1,396 stars in a single day. The project, described as making "your AI agent think like the laziest senior dev in the room," has become one of the fastest-growing repositories on GitHub. The core philosophy - "the best code is the code you never wrote" - resonates with developers tired of AI agents that generate unnecessary complexity. Built in JavaScript with 6,329 forks, ponytail has attracted contributions from the Claude team itself, with @claude listed among the top contributors. The repo's meteoric rise suggests it fills a real gap in how AI coding agents approach code generation.
Hebbian Robotics (YC S26): Scalable Robotics Data Pipelines
Hebbian Robotics launched from Y Combinator's Summer 2026 batch with hflow, a framework for building scalable robotics data pipelines. The project appeared on Hacker News with 43 points, targeting the growing need for infrastructure to collect, process, and manage the massive datasets required for training robotic systems. As robotics companies scale from prototype to production, the data pipeline challenge - collecting sensor data, managing ground truth annotations, and versioning training datasets - has become a critical bottleneck. Hebbian's approach focuses on making these pipelines reproducible and scalable across different robot configurations.
EVA-Client: Unified Framework for Real Robot Deployment and Evaluation
EVA-Client has trending on Trendshift as the #2 project, providing a unified framework for deployment, evaluation, and data collection on real robots. The project addresses a persistent challenge in robotics: the gap between simulation and real-world deployment. EVA-Client standardizes the interface for deploying models to physical robots, collecting performance data, and iterating on improvements. The framework supports multiple robot platforms and provides tools for systematic evaluation across different environments and conditions.
Archify: Diagram Generation from Natural Language (35K Stars)
Archify has reached 35,000 GitHub stars as an AI-powered tool that generates architecture, workflow, sequence, data-flow, and lifecycle diagrams from natural language descriptions. The project creates self-contained HTML diagrams with motion and crisp export capabilities. It has become a go-to tool for developers and architects who need to quickly visualize system designs without manually drawing in diagramming tools. The project's growth reflects the broader trend of AI tools replacing manual creative and documentation tasks.
Scientific Agent Skills: 165 Validated Skills for AI Scientists (39K Stars)
Scientific Agent Skills has reached 39,000 GitHub stars, offering 165 validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. The project turns any AI agent into an AI scientist, with skills compatible with Cursor, Claude Code, Codex, and other AI coding clients. Used by 190,000+ scientists worldwide, it represents the growing intersection of AI agents and scientific research. The skills cover everything from protein structure prediction to chemical property calculation, making AI-powered research more accessible to scientists without deep ML expertise.
Why it matters. This week's deals - totaling over $85 billion - confirm that AI infrastructure is consolidating into the hands of a few giants. Nvidia buying Hugging Face gives the chipmaker direct control over the open-source model ecosystem. Stripe acquiring OpenRouter puts payments infrastructure in charge of AI model routing. Anthropic's $65 billion revenue run rate shows that the AI economy is no longer speculative - it's generating more revenue than most Fortune 500 companies. Meanwhile, Apple's supply constraints reveal that the hardware layer can't keep up with AI demand, and Groq's pivot from chips to cloud shows that even well-funded startups must adapt to Nvidia's dominance. For developers, the message is clear: the tools you build on are being acquired, and the infrastructure layer is being reshaped by forces measured in tens of billions.