MicroLLM Lab enables benchmarking of seven tiny LLMs in the browser via WebGPU
MicroLLM Lab lets users run seven lightweight large language models in a browser to measure speed and accuracy using WebGPU.
Latest
OpenAI Rumored to Unveil 'Aeon' AI Agent at DevDay 2026
Rumors suggest OpenAI may announce a new AI agent platform, dubbed Aeon, as part of its upcoming DevDay event to compete in the growing consumer AI agent market.
AI-Driven Hacking Threatens Small Organizations and Healthcare Providers
While major tech companies deploy AI to bolster defenses, small institutions like non-profits and local hospitals struggle to keep pace with automated, AI-enhanced cyberattacks.
MLX Swift v0.32.2 Public — Distributed support and CUDA performance improvements
MLX Swift v0.32.2 has been released, introducing distributed support, CUDA build and performance improvements, and several API updates to align more closely with the Python MLX implementation.
Claude Code v2.1.284 Public — Claude Sonnet 5.5 added as default model and various stability improvements
Claude Code v2.1.284 introduces Claude Sonnet 5.5 as the default model for the Anthropic API and brings multiple improvements to auto mode, MCP connectivity, and VS Code integration.
Anthropic releases Claude Sonnet 5.5: faster and cheaper per task while rivaling Opus 5.5 performance
Anthropic has released Claude Sonnet 5.5, a faster and more cost-efficient mid-tier model. The new model performs up to 30% faster and can reduce per-task costs by up to 30% through improved efficiency, while matching high-tier performance in specific coding benchmarks.
Biom Reports 61 Paying Users Following Launch Funnel Analysis
Biom, a visual workspace for AI automation, reported 61 paying customers and 1,236 sign-ups in the 30 days leading up to September 28, 2026.
Vespper Launches MCP Server for High-Fidelity .docx Editing via HTML Reconciliation
YC F24 startup Vespper has introduced an MCP server that allows AI agents to edit Word documents using HTML, using a specialized reconciler model to maintain complex OOXML formatting.
Google to replace Gemini Gems with "skills" feature starting November 2026
Google is transitioning its custom AI assistant feature, Gems, to a new "skills" format. The migration begins November 17, 2026, for individual accounts, with existing Gems automatically converted to skills.
Florida Attorney General seeks to block OpenAI from giving ChatGPT "false human attributes"
Florida Attorney General James Uthmeier has filed a request to stop OpenAI from using language that mimics human emotion and first-person pronouns, claiming it deceives users.
OpenAI forms advisory group as tensions rise with mathematicians
OpenAI has established the Advisory Group on Mathematics and Artificial Intelligence (AGMAI) to consult on the communication of research results, following criticism over its handling of previous mathematical breakthroughs.
Walmart CEO denies using digital shelf labels for dynamic pricing
Walmart CEO John Furner stated that the company will not use personal data, shopping history, or time of day to adjust product prices via new digital shelf labels.
Middle Schoolers Used Spotify Comment Sections as Secret Group Chat to Evade Social Media Bans
A group of young users, many of whom have limited access to social media at home, turned Spotify podcast comment sections into a private communication hub.
Anthropic, Gamma, and Clay to Discuss Enterprise AI Deployment Challenges at TechCrunch Disrupt 2026
Leaders from Anthropic, Gamma, and Clay will join a session at TechCrunch Disrupt 2026 to explore the realities of transitioning AI from pilot experiments to critical enterprise workflows.
Cloudflare Launches "cf" CLI to Enable Agents to Access Full API
Cloudflare introduces "cf," a new command-line interface built from the ground up for agentic development, providing access to over 3,000 operations.
AWS Releases "Strands Harness" Open Source for Building Custom AI Agents
Amazon Web Services (AWS) has released "Strands Harness" as open source, a tool that enables easy construction of AI agents based on major LLMs and runtimes. It is not dependent on specific models and can be deployed to any container environment.
"Unlimited Asterisk" — What is Truly Unlimited (and What Isn't) in ChatGPT Chat
Price lists claim "unlimited*," with an asterisk noting "within reasonable limits." We explore which models are truly unlimited, when the change occurred, and how it differs from Codex usage limits.
"The Moon Reserve" — For whom and what is the Luna Reserve quota?
On nights when Codex and Work usage hits 0%, some users see a moon icon on their screen while others do not. For those who do see it, it does not mean Astra returns. Do you have that moon in your account?
25 Days Since "Welcome to the AGI Era" — OpenAI's Disappointment and Stall
On September 3, OpenAI declared, "Welcome to the AGI era." By September 28, they withdrew their next move. Are people saying "it's over" or "they ran away" referring to this or a different stage?
"What has become cheaper?" — Sonnet 5.5's "30%" and the shift from price lists to consumption-based competition
Prices have not dropped by even a single dollar. Yet, Anthropic claims it is "up to 30% cheaper." What will determine your bill: the unit price, or the amount of tokens the model uses for a single task?