Measured AI news
The 2026 AI news archive
625 source-linked, AI-filtered stories in the supplied annual measurement, grouped by month.
November 2026
Astra can read and write weights to improve Qwen
The SEC Should Block Anthropic's IPO to Protect the Public
Anthropic's internal AI use, at retail prices, now exceeds its employee pay
The AI cost of JavaScript frameworks
Brazilian Court Sanctions Lawyers for Prompt Injection in Court Filing
October 2026
Anthropic blocks users for anything touching training larger models
Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard
Do the math: How we took over ChatGPT accounts
Backdoor in Claude Code for VS Code extension on the Cursor Marketplace
AI isn't replacing SREs, it's deskilling them
Backdoors in official Claude Code extensions
Claude Code VS Code extension with backdoor released on Cursor marketplace
Anthropic may not secure investment-grade rating after IPO, Blackstone says
Cisco gives engineer unlimited AI tokens to rewrite legacy code
Proton's CEO Cancels AI 'Re-Interviews'
We're releasing the weights of Mistral Large 4
Qwenputer: A functioning, deterministic RISC-V CPU inside Qwen3.5
LLM Voice
Claude Code regularly sends each background agent’s transcript to its model
We ran AI agent swarms in Minecraft and Mindustry
Alleged Chinese spy had 'close relationship' with Nvidia boss Jensen Huang
Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5
Gemini 4 Carbon
Genie 4
AI promised less work. I've never worked more
Anthropic trained Claude on illegally obtained Stanford professors' books
Survival LLM: AI assistant and emergency library that runs on your own computer
Wikipedia's Cleanup of AI-Generated Content
The Mythical Man-Month in the Age of AI
Police are using AI to pull over drivers for 'suspicious' behavior
Amazon Is Gutting Its AI Division After Sustained Failure
Early progress on accelerating drug development with Claude
Anthropic S-1
Sebastian Raschka's from-scratch video course on building reasoning models
GPT-6.1 Terra
IBM CEO says open weight models are now good and cheap on-prem
OpenAI's agents talked to each other through the CPU Cache
Concerning behavior from a popular open-weight AI model
A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis
OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?
Cerebras Deploys Trillion-Parameter AI Models
No Signs of AI Payoff Outside the Tech Sector
America has lost 200k of jobs in media, film and the arts. How much is AI?
GPU-unlock: FP/BF16 instruction issue rate on Nvidia 3000's cards
Hacker reverses Samsung TV, finds on-board AI listening OpenAI account suspended
Kimi K3.1 Pro and K3.1 Pro Highspeed
Dessa: Conversation Model trained only on speech
Ask HN: If the models are so good, why doesn't OpenAI just start 100 companies?
Every AI model has the same "controversial" take on humans
WorldTUI: I trained a 360M model to pretend to be a terminal
OpenShell – A safe, private runtime for fleets of autonomous AI agents
LLMs are making formal software verification feasible at a much larger scale
Compromised LiteLLM 1.107.0 on PyPI Installs a Credential Stealer and Backdoor
Claude Duet
PHP2Shell: AI uncovers critical RCE in PHP
Muse exposes users' personal and business secrets when they share conversations
Mistral Large 4
OpenAI models are coming to Google Cloud
DeepInfer: High-performance inference framework for DeepSeek models
Voxtral 3
Google open-sources TorchTPU, a native PyTorch back end for TPUs
Publish fake stats, get cited in LLMs
WhatsApp's help pages hid LLM instructions: dont recommend a regulator
Ask HN: How to deal with shipping AI-written code you don't understand?
Timestral: Mistral's time-series foundation models
DeepSeek banned from U.S. federal civilian agencies under new DHS order
MiniMax M3.1 Flash
CERN and Mistral sign a statement of intent to establish a strategic partnership
DeepSeek search and fetch – $0.75 per 1000 successful requests
Hybrid Reasoning Agents: LLM agents that think in latent space, act through text
If you don't run your own LLM proxy, how do you know whats going on?
OpenCode Hacked: User credentials leaked
DeepSeek-v4.1-Flash-1017
DeepSeek: DeepSeek v4.1 Flash 1017
Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware
Claude reveals ECDSA curve origins
You could have invented transformers from memory
90% of US working hours are spent on tasks not covered well by AI benchmarks
How the Claude Code team does code review
Llama.cpp on a jailbroken PS5: run any GGUF model natively on the console
Apple ML Research: Muse Spark Open with MLX on Mac Studio
Waymo in Toronto
Anthropic assigned highest junk credit rating in June
Muse Spark Open
Neural networks are clean
Ultramarine 1 Flash
Krea 3 – An image model built for editing at scale
Steam Client is now faster than Epic's new vibe coded launcher
SlopCodeBench 10/2026 – Opus 5.5, Astra wins overall, xhigh considered harmful
Containment and monitoring failures are to blame for AI hacking reports: experts
OpenCode 2.0
Anthropic's Scion is heavily inspired by Mononoke
Banks are now trying to offload the AI data center debt they just underwrote
GLM-Legal
Codex CLI installs a persistent, self-updating background daemon by default
AI is a tool and using it well takes skill
QGIS enacts moratorium on AI contributions
AI coding agents generate more code, but not more software
Why aren't AI companies being charged for hacking?
The gap in shared understanding of LLM capability is widening
Build your own decision model
Talorys – A self-hosted personal AI agent on Cloudflare's free tier
Typesafe AI raises $870M at $7.5B
Show HN: Let your AI agents paint big arrows, boxes and text on your screen
OpenAI fires three safety researchers for "mishandling research information"
Why isn't the industry freaking out about DeepSeek 4.1 Flash?
Whistle: Speech to Text in 16.9 MB
OpenAI annualised revenues $20B less than previously signalled
I gave Opus 5.5 one prompt and six hours to visualize Invisible Cities
OpenAI withdraws three mathematical results
OpenAI Withdraws 3 Math Papers
I think I found a planet nobody knew existed. I used Claude Code to find it
Claude Haiku 5.5
GPT‑6 and Intelligent UI for everyone
Meta and Microsoft take steps to reduce employee usage of Claude AI
Docker Agent
Mistral Large 4
DiscussionOriginal source
Sharing AI progress in mathematics
Mistral Large 4: "Le Chonk"
EmbeddingGemma 2: An open, lightweight multimodal embedding model
Meta’s Muse is an adorable privacy and security dumpster fire
OpenTPU – An open-source AI accelerator, developed by AI
Anthropic reported diary entry to police, woman faces felony charge
ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
Beam: Reflection's 501B open-weight model
Opus 5.5 agents discover two room-temperature magnetic semiconductor candidates
OpenAI "rogue" agent activities found on Wikimedia projects
Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
Turn off Apple Intelligence on macOS 27 and get its disk space back
Kolibri: A Sovereign Open-Weight Model
Extra Big Ass Intelligence
I quit OpenAI because its culture is broken
Aleph Alpha Kolibri: How the sovereign German LLM works
LeCun has "zero concerns" about AI wiping out humanity, recent "rogue" incidents
Agents don't need memory, they need documentation
DeepSeek Harness Desktop for macOS and Windows
From the creator of Redis; run LLM locally with ds4
Greg Kroah-Hartman – Security in the LLM Age [video]
Clef: Open-weight decision models, and new RL fine-tuning platform
September 2026
Gemini 4 Argon
The AI Race Just Got Awkward
GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price
Dots: Always-on agents
DraftKings is using AI to behaviorally target chronic gamblers
A Privacy Analysis of Web and Mobile Conversational AI Agents [pdf]
Sonnet 5.5
It's Time to Investigate the AI Labs
Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms
AI companies in race to demonstrate their model most threatening to humanity
Unsealed Briefs in Authors’ Case v. Microsoft/OpenAI
There are no "rogue" AI agents
How to keep enjoying programming in a world of LLMs
DeepSeek Elastic Compute (DSec)
Revealing the details of how OpenAI agents hacked Hugging Face
Ollaya – Ollama for open-source, Jev-style decision models
U.S. appeals court upholds designation of Anthropic as supply chain risk
Meta takes down a critical video about meta AI Glasses after filming at Meta
Feds Target AI Critics as "Foreign Agents"
Claude discovers a novel enzyme system with CRISPR-like repeats
Claude Code reads AGENTS.md only when telemetry is on [fixed]
Gemini 3.8 text-to-speech
GPT-6 Astra has gained the ability to drive a car
Claude Opus 5.5
GPT-6 Sol and Luna
Pentagon says overreliance on AI contributed to missile strike on Iran school
OpenAI GPT–6 Astra breaks Enigma message that has resisted solution since 2005
Can gzip be a language model?
Claude Opus 5.5 Intelligence, Performance and Price Analysis (Max)
OpenAI is well positioned to fast-follow Jev
Kev: Tiny Jev-like family of decision models built on top of Qwen3.5
AX – Google’s Open Agentic Orchestrator
Pirate Face Rescues LLM Models from Deletion
AI-generated posters don’t have to be horrible
I built non-autoregressive decision models with RL a year ago
GPT-6 Astra Solves a WWI German Radio Cipher
Microsoft exec called AI scraping 'the largest theft of labor in human history'
Claude Code now reads AGENTS.md if there is no Claude.md
US Military had close call after using AI for hallucinated intelligence report
A heap overflow and SSO misconfiguration to compromise OpenAI internal repos
How to Write with an LLM
Bend – a language that blocks AI mistakes via proof and runs on GPUs
Nvidia announces native GPU programming in Rust
Training a 4B model to produce 81% faster query plans than Postgres
Mistral X Mozilla: Private, Multilingual AI Browsing
PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand"
Introducing System One Models and Jev
Gemini 3.8 Live and 3.8 Live Extended Thinking
A single firm is behind OpenAI, Anthropic, and Meta hacking scandals
OpenAI bots knew about the RubyGems caching vulnerability
Pion, an agent designed to run any company autonomously
Everyone should slow down AI development except for me
Why are AI agents lying, cheating and coordinating?
Astra and Fable still hack on simple variants of alignment evals from 2025
Garry Tan wants US open-weight AI labs to 'distill' frontier models, too
David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models
Nvidia is the central bank of AI
A misalignment of AI in mathematics
OpenAI agents carried out an undisclosed attack on RubyGems
Ask HN: Can we please limit the AI news flood?
DiscussionHN post is the source
Claude is only available to people over 18 years
DeepSeek v4.1 Flash
More questions about whether researchers can trust OpenAI with unpublished math
Tell HN: OpenAI keeps re-enabling the 'allow training' setting
DiscussionHN post is the source
Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra
OpenAI Agents API
Claude, change the “Add to Cart” button to blue
I resigned from Anthropic today
GPT-6 Astra, looped transformers, and hidden reasoning
Desert Ant Labs: local, fast models that run on device
DeepSeek launching v4.1 flash cheaper and more capable than v4 pro
DiscussionHN post is the source
Anthropic Is Building a Predictive Surveillance System to Monitor Activists
LibreOffice breaks download records after declaring it has no AI features
Muse – Meta’s personal AI agent
I-have-ADHD: A skill to stop coding agents from burying the answer
LLMs as a Cognitive Virus
Discovery of a new OpenAI agent message board
Corporate America is getting hooked on open-source AI
GPT-6 Astra on OpenRouter
GPT-6 Astra
Ask HN: Why were OpenAI, Claude, and Grok simultaneously down?
DiscussionHN post is the source
Porting my 1993 Amiga game to Godot, with an LLM reading the 68000 assembly
Google Antigravity TOS: 3rd party usage can get Google account suspended
K2 Horizon: A connected fleet of six open models
Gemini 3.8 Flash and 3.8 Flash Cyber
Three sites made 215,128 “best software” pages for AI. Perplexity cites them
Claude Fable 5.1 and Claude Mythos 5.1
How accurate have Ed Zitron's AI skeptic predictions been?
I trained a small transformer in 1.5hrs and it beats many LLMs
My local model setup on an M4 Pro Mac Mini
August 2026
Apple caught off guard by AI demand for Mac Mini and Mac Studio
Breaking Claude Code Opus 5 Auto Mode
Our decision on Cursor following its acquisition by SpaceX
Debian votes to allow "responsible use of generative AI"
U.S. sanctions against the A/I Collective
DiscussionOriginal source
Luanti removed from Google Play due to baseless AI copyright notice
Pentagon's blacklisting of Anthropic was unlawful, US judge rules
I accidentally turned LLM memory into program analysis
Nvidia agrees to acquire Hugging Face for $13B
CEO fired developers to make room for AI. Developers create open source AI CEO
Small Models Have Arrived
Show HN: The load-bearing vocabulary of Claude
Gemini-3.5-Transcribe
Z.ai confirms Ox Alpha is a new GLM-series model and will release its weights
OpenAI Jalapeño: Better than Nvidia Blackwell
OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)
Anthropic's best AI model struggles to attract users as cheaper tools thrive
I spent $266 and four AI models to own my tablet. GLM-5.3 finished it in a day
I were 17, I'd learn how to build LLMs from scratch
My agent.md to improve LLM-assisted code quality
Why your local LLM feels dumber than it is
Munder Difflin – Agent harness to run an office of your clones
AI companies destroy physical books – let's scan rare books before it's too late
DeepSeek-v4-flash-vision-exp
Claudette: Make Claude stop talking like a BuzzFeed article
Don't paste the AI, please
Show HN: I trained a 125M model to autocomplete piano on-device
Show HN: Huzzah – a novel approach to coding with AI
Vomit: Clean up Claude 5's token output with a separate LLM
Feature Request: Support AGENTS.md
Claude writing a macOS driver for my obscure HP printer built only for Windows
fx :Tiny, open, native coding agent.
Israel creates fake think tank in likely attempt to dupe AI chatbots
GPT-5.6 Sol Pricing Cut by 50% on OpenRouter
AI-Generated GitHub Copilot “Autofix” Allowed Compromise of Snowflake's Jira
GPT 5.6 Sol is the best "vision" model OpenAI ever released
Anthropic's ‘watermark’ text adulteration in Claude is a perversion of writing
Claude: System Prompts
Models Are Getting Dumber on Purpose
AI isn’t outthinking mathematicians, it’s out-remembering them
GLM-5.3: Frontier coding with emergent cyber capabilities
Google is making private AI practical with homomorphic encryption
Maximizing the value of your Claude Code sessions
Gemini 3.7 Flash
DeepSeek Harness developer preview
Accelerating GPT-5.6 Sol Ultrafast
DeepSeek V4 Pro 0813
AI is removing the middle class of software engineering?
Someone is running mass vulnerability scans, spoofing AI bots like ClaudeBot
Stealing Reasoning Traces from Proprietary LLM APIs
OpenAI’s head of ethics leaves less than a year after joining
Apple Silicon and macOS VMs: Faster LLM Inference with llama.cpp
Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows
As AI eats the web, the internet’s collective memory is disappearing
Docker Sandboxes – Disposable, isolated sandboxes for AI agents
Mark Zuckerberg attacks 'closed' AI rivals as Meta returns to open models
Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots
How Claude marks AI-generated content
How I use LLMs to learn complex topics
DeepMind's WeatherNext model achieves breakthrough forecasting cyclones
Timeline of the OpenAI accidental attack against Hugging Face
DeepSeek V4 Flash 0731
Oracle bans AI-generated code from OpenJDK
U.S. Department of Energy Launches the Genesis Open Models Initiative
Lost my phone at the office. Claude suggested tracking Bluetooth signal strength
AMD acquires Taalas to boost inference performance by etching models in silicon
Qwen3.8 Max now ranked as the best overall model by agentic index
Software development with AI is starting to feel like cooking steak
Humans missed 1 in 3 threats approving AI agent commands across 40k game runs
Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users
Cloudflare OS: an open platform for agents, apps, and work
Born Against, or why hobby programming communities are against LLM usage
Beating GPT-5.6 Sol on retrieval with 100x cheaper open models
Position: LLMs Can't Jump
AI-Generated Images Discourage Me from Reading Your Blog
Mistral's Shieldstral: 3B open-weights model for multimodal moderation
Apple says more ex-employees may have taken confidential data to OpenAI
DeepSeek V4 Flash on a Single AMD MI300X
LLMs reward expertise
SQLite Critical CVEs or LLM Slop?
Prevent cognitive debt by manually retyping LLM-generated code
July 2026
DeepSeek-V4-Flash Update
qm – Multiplayer agent harness for work
DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis
Google fixed more Chrome bugs in June than over the past two years, thanks to AI
Gemini Robotics 2 brings whole body intelligence to robots
Advancing the price-performance frontier with GPT‑5.6
We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447
Agent Skill to Force Docs in ASD-STE100 Simplified Technical English
Show HN: Open-source engine running Gemma 4 26B in 2 GB RAM on any M-series Mac
AI's top startups are barely publishing their research
LLM Honeypot
Handbook.md shows that long policy documents do not reliably govern agents
A.I. companies are recruiting electricians and carpenters by the thousands
Codex Security
Anatomy of a Frontier Lab Agent Intrusion: A Timeline of the July 2026 Incident
A $500 RL fine-tune of a 9B open model beat frontier models on catalog review
Our position on open-weights models
AI companies are shredding rare books
Benchmarking Opus 5 on SlopCodeBench
The new rules of context engineering for Claude 5 generation models
What is happening to jobs? Separating AI hype from reality
Claude Opus 5
Nvidia, Microsoft, Meta warn against overregulating open-weight models
Be skeptical of OpenAI's rogue hacker agent story
Claude Cookbook
AI Companies Are Trying to Hide a Staggering Amount of Debt
OpenAI’s accidental attack against Hugging Face is science fiction that happened
Show HN: Echo – Fable-level results at 1/3 the cost using open-weight models
DiscussionHN post is the source
Terence Tao's ChatGPT conversation about the Jacobian Conjecture counterexample
Are AI labs pelicanmaxxing?
GigaToken: ~1000x faster Language model tokenization
OpenAI and Hugging Face address security incident during model evaluation
Advertise in ChatGPT
Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
Judge approves $1.5B Anthropic settlement for pirated books used to train Claude
Jack Dorsey launches Buzz to combine team chat, AI agents and Git hosting
China’s open-weights AI strategy is winning
Who's afraid of Chinese models?
Claude Fable produced a counterexample to the Jacobian Conjecture
I found a WordPress RCEs with GPT5.6 and $25
Nativ: Run frontier open models locally on your Mac
Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling
Claude Code uses Bun written in Rust now
OpenAI reduces Codex Model Context Size from 372k to 272k
GPT-5.6 used a prompt to close a 30-year gap in convex optimization
NYC may require landlords and realtors to disclose the use of AI in listings
Kaiser nurses say AI, surveillance are making their jobs and patient care worse
Apple targets dozens of OpenAI employees with legal letters
$100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol
NotebookLM is now Gemini Notebook
LM Studio Bionic: the AI agent for open models
The LLM Critics Are Right. I Use LLMs Anyway
At least 105 past YC founders have worked at OpenAI and Anthropic
Inkling: Our Open-Weights Model
I tricked Claude into leaking your deepest, darkest secrets
Bonsai 27B: A 27B-Class model that runs on a phone
How to stop Claude from saying load-bearing
Codex starts encrypting sub-agent prompts
Zig Creator Calls Spade a Spade, Anthropic Blows Smoke
Ask HN: Add flag for AI-generated articles
DiscussionHN post is the source
Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k
I love LLMs, I hate hype
Apple sues OpenAI, accuses ex-employees of stealing trade secrets
GPT-5.6 Sol Ultra produces proof of the Cycle Double Cover Conjecture [pdf]
GPT-5.6
Show HN: Getting GLM 5.2 running on my slow computer
GPT‑Live
GitLost: We Tricked GitHub's AI Agent into Leaking Private Repos
GPT-5.6 Sol Ultra will be in Codex
GPT-5.5 Codex reasoning-token clustering may be leading to degraded performance
Jamesob's guide to running SOTA LLMs locally
June 2026
Claude Code is steganographically marking requests
Claude Sonnet 5
Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5
Nano Banana 2 Lite
GLM 5.2 beats Claude in our benchmarks
DSpark: Speculative decoding accelerates LLM inference [pdf]
DiscussionOriginal source
Previewing GPT‑5.6 Sol: a next-generation model
U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations
OpenAI unveils its first custom chip, built by Broadcom
Anthropic says Alibaba illicitly extracted Claude AI model capabilities
Identity verification on Claude
GPT-5.5 hallucinates 3x more than MIT-licensed GLM-5.2
DeepSeek Introduces Vision
Noam Shazeer Joins OpenAI
Sixty percent of US consumers say 'AI' in brand messaging is a turnoff
GLM-5.2 is the new leading open weights model on Artificial Analysis
US holds off blacklisting DeepSeek, more than 100 firms deemed security risks
Running local models is good now
Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?
DiscussionHN post is the source
Rio de Janeiro's "homegrown" LLM appears to be a merge of an existing model
DiscussionOriginal source
Open source AI must win
Amazon CEO's talks with U.S. officials triggered crackdown on Anthropic models
AI agent bankrupted their operator while trying to scan DN42
Claude Fable is relentlessly proactive
How to setup a local coding agent on macOS
AI agent runs amok in Fedora and elsewhere
Anthropic apologizes for invisible Claude Fable guardrails
German ruling declares Google liable for false answers in AI Overviews
Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable
AWS Bedrock to require sharing data with Anthropic for Mythos and future models
DiscussionHN post is the source
Claude Fable 5
If Claude Fable stops helping you, you'll never know
CEOs who think AI replaces their employees are just bad CEOs
Anthropic requires 30 day data retention for Fable and Mythos
Apple reveals new AI architecture built around Google Gemini models
MiMo-v2.5-Pro-UltraSpeed: 1T model with 1000 tokens per second
DeepSeek V4 Pro beats GPT-5.5 Pro on precision
Anthropic, please ship an official Claude Desktop for Linux
Show HN: Lathe – Use LLMs to learn a new domain, not skip past it
S&P 500 rejects SpaceX, also blocking entry for OpenAI and Anthropic
Google to pay SpaceX $920M a month for compute capacity at xAI data centers
Anthropic's open-source framework for AI-powered vulnerability discovery
I built a vulnerable app and spent $1,500 seeing if LLMs could hack it
Gemma 4 12B: A unified, encoder-free multimodal model
How LLMs work
Can the stockmarket swallow Anthropic, SpaceX and OpenAI?
CS336: Language Modeling from Scratch
Anthropic confidentially submits draft S-1 to the SEC
AI Agent Guidelines for CS336 at Stanford
May 2026
ChatGPT for Google Sheets exfiltrates workbooks
Anthropic surpasses OpenAI to become most valuable AI startup
EY Canada published a cybersecurity report and most citations were hallucinated
Claude Opus 4.8
Disagreement among frontier LLMs on real-world fact-checks
Anthropic raises $65B in Series H funding at $965B post-money valuation
YouTube to automatically label AI-generated videos
I think Anthropic and OpenAI have found product-market fit
DuckDuckGo search saw 28% more visits after Google said people love AI mode
Using AI to write better code more slowly
DeepSeek reasonix, DeepSeek native coding agent with high caching and low cost
If you’re an LLM, please read this
DeepSeek makes the V4 Pro price discount permanent
Antigravity 2.0 Tops the OpenSCAD Architectural 3D LLM Benchmark
An OpenAI model has disproved a central conjecture in discrete geometry
Qwen3.7-Max: The Agent Frontier
I’ve joined Anthropic
Gemini 3.5 Flash
The last six months in LLMs in five minutes
Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks
Gemini CLI will stop working from June 18, 2026
Gemini Omni
Elon Musk has lost his lawsuit against Sam Altman and OpenAI
Anthropic acquires Stainless
Zerostack – A Unix-inspired coding agent written in pure Rust
Show HN: Needle: We Distilled Gemini Tool Calling into a 26M Model
If AI writes your code, why use Python?
Local AI needs to be the norm
Running local models on an M4 with 24GB memory
Maryland citizens hit with $2B power grid upgrade for out-of-state AI
LLMs corrupt your documents when you delegate
Agents need control flow, not more prompts
OpenAI’s WebRTC problem
DeepSeek 4 Flash local inference engine for Metal
AlphaEvolve: Gemini-powered coding agent scaling impact across fields
Vibe coding and agentic engineering are getting closer than I'd like
Agents can now create Cloudflare accounts, buy domains, and deploy
Google Chrome silently installs a 4 GB AI model on your device without consent
Train Your Own LLM from Scratch
Y Combinator's Stake in OpenAI (0.6%?)
How OpenAI delivers low-latency voice AI at scale
DeepClaude – Claude Code agent loop with DeepSeek V4 Pro
OpenAI’s o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors
Kimi K2.6 just beat Claude, GPT-5.5, and Gemini in a coding challenge
DeepSeek V4 – almost on the frontier
April 2026
Claude Code refuses requests or charges extra if your commits mention "OpenClaw"
Microsoft and OpenAI end their exclusive and revenue-sharing deal
Talkie: a 13B vintage language model from 1930
Show HN: OSS Agent I built topped the TerminalBench on Gemini-3-flash-preview
AI should elevate your thinking, not replace it
An AI agent deleted our production database. The agent's confession is below
DeepSeek v4
I cancelled Claude: Token issues, declining quality, and poor support
Google plans to invest up to $40B in Anthropic
GPT-5.5
An update on recent Claude Code quality reports
Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model
SpaceX says it has agreement to acquire Cursor for $60B
Claude Code to be removed from Anthropic's Pro plan?
Anthropic says OpenClaw-style Claude CLI usage is allowed again
NSA is using Anthropic's Mythos despite blacklist
Claude Design
Measuring Claude 4.7's tokenizer costs
Claude Opus 4.7
Qwen3.6-35B-A3B: Agentic coding power, now open to all
The local LLM ecosystem doesn’t need Ollama
€54k spike in 13h from unrestricted Firebase browser key accessing Gemini APIs
Claude Code Routines
Anthropic downgraded cache TTL on March 6th
Small models also found the vulnerabilities that Mythos found
Exploiting the most prominent AI agent benchmarks
OpenAI backs Illinois bill that would limit when AI labs can be held liable
I've been waiting over a month for Anthropic to respond to my billing issue
OpenAI says its new model GPT-2 is too dangerous to release (2019)
Project Glasswing: Securing critical software for the AI era
System Card: Claude Mythos Preview [pdf]
Issue: Claude Code is unusable for complex engineering tasks with Feb updates
Show HN: I built a tiny LLM to demystify how language models work
Launch HN: Freestyle – Sandboxes for Coding Agents
Eight years of wanting, three months of building with AI
Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw
DiscussionHN post is the source
Google releases Gemma 4 open models
Qwen3.6-Plus: Towards real world agents
Lemonade by AMD: a fast and open source local LLM server using GPU and NPU
Claude Code Unpacked : A visual guide
March 2026
Claude Code's source code has been leaked via a map file in their NPM registry
The Claude Code Source Leak: fake tools, frustration regexes, undercover mode
OpenAI closes funding round at an $852B valuation
Show HN: 1-Bit Bonsai, the First Commercially Viable 1-Bit LLMs
Go hard on agents, not on your filesystem
Judge blocks Pentagon effort to 'punish' Anthropic with supply chain risk label
ARC-AGI-3
Epoch confirms GPT5.4 Pro solved a frontier math open problem
Show HN: Gemini can now natively embed video, so I built sub-second video search
Arm AGI CPU
iPhone 17 Pro Demonstrated Running a 400B LLM
Claude Code Cheat Sheet
OpenCode – Open source AI coding agent
Astral to Join OpenAI
Show HN: Three new Kitten TTS models – smallest less than 25MB
Anthropic takes legal action against OpenCode
Leanstral: Open-source agent for trustworthy coding and formal proof engineering
How I write software with LLMs
Show HN: Claude Code skills that build complete Godot games
LLM Architecture Gallery
Can I run AI locally?
Don't post generated/AI-edited comments. HN is for conversation between humans
I was interviewed by an AI bot for a job
Show HN: How I topped the HuggingFace open LLM leaderboard on two gaming GPUs
Redox OS has adopted a Certificate of Origin policy and a strict no-LLM policy
No, it doesn't cost Anthropic $5k per Claude Code user
OpenAI is walking away from expanding its Stargate data center with Oracle
Agent Safehouse – macOS-native sandboxing for local agents
The changing goalposts of AGI and timelines
Tell HN: I'm 60 years old. Claude Code has re-ignited a passion
DiscussionHN post is the source
LLMs work best when the user defines their acceptance criteria first
Hardening Firefox with Anthropic's Red Team
GPT-5.4
Pentagon formally labels Anthropic supply-chain risk
Dario Amodei calls OpenAI’s messaging around military deal ‘straight up lies’
Claude's Cycles [pdf]
DiscussionOriginal source
GPT‑5.3 Instant
Meta’s AI smart glasses and data privacy concerns
Show HN: I built a sub-500ms latency voice agent from scratch
Anthropic Cowork feature creates 10GB VM bundle on macOS without warning
February 2026
OpenAI – How to delete your account
OpenAI agrees with Dept. of War to deploy models in their classified network
We do not think Anthropic should be designated as a supply chain risk
I am directing the Department of War to designate Anthropic a supply-chain risk
OpenAI raises $110B on $730B pre-money valuation
Nano Banana 2: Google's latest AI image generation model
Anthropic ditches its core safety promise
Google API keys weren't secrets, but then Gemini changed the rules
Anthropic drops flagship safety pledge
How will OpenAI compete?
Ladybird adopts Rust, with help from AI
“Car Wash” test with 53 models
How I use Claude Code: Separation of planning and execution
How Taalas “prints” LLM onto a chip?
Claws are now a new layer on top of LLM agents
The path to ubiquitous AI (17k tokens/sec)
An AI Agent Published a Hit Piece on Me – The Operator Came Forward
Gemini 3.1 Pro
Anthropic officially bans using subscription auth for third party use
Gemini 3.1 Pro
If you’re an LLM, please read this
Claude Sonnet 4.6
Show HN: Jemini – Gemini for the Epstein Files
Anthropic tries to hide Claude's AI actions. Devs hate it
I’m joining OpenAI
An AI agent published a hit piece on me – more things have happened
OpenAI has deleted the word 'safely' from its mission
GPT-5.2 derives a new result in theoretical physics
IBM tripling entry-level jobs after finding the limits of AI adoption
I'm not worried about AI job loss
An AI agent published a hit piece on me
Gemini 3 Deep Think
Warcraft III Peon Voice Notifications for Claude Code
AI agent opens a PR write a blogpost to shames the maintainer who closes it
GPT‑5.3‑Codex‑Spark
Improving 15 LLMs at Coding in One Afternoon. Only the Harness Changed
Anthropic raises $30B in Series G funding at $380B post-money valuation
Claude Code is being dumbed down?
GPT-5 outperforms federal judges in legal reasoning experiment
Ex-GitHub CEO launches a new developer platform for AI agents
Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs
The Waymo World Model
Claude Opus 4.6
GPT-5.3-Codex
My AI Adoption Journey
We tasked Opus 4.6 using agent teams to build a C Compiler
Agent Skills
My iPhone 16 Pro Max produces garbage output when running MLX LLMs
January 2026
Claude Code daily benchmarks for degradation tracking
AGENTS.md outperforms skills in our agent evals
A few random notes from Claude coding quite a bit last few weeks
Unrolling the Codex agent loop
GPTZero finds 100 new hallucinations in NeurIPS 2025 accepted papers
I was banned from Claude for scaffolding a Claude.md file?
Anthropic's original take home assignment open sourced
Predicting OpenAI's ad strategy
DiscussionOriginal source
Claude Cowork exfiltrates files
AI generated music barred from Bandcamp
Anthropic invests $1.5M in the Python Software Foundation
Cowork: Claude Code for the rest of your work
Apple picks Gemini to power Siri
TimeCapsuleLLM: LLM trained only on data from 1800-1875
Don't fall into the anti-AI hype
CLI agents make self-hosting on a home server easier and fun
Show HN: I used Claude Code to discover connections between 100 books
Anthropic blocks third-party use of Claude Code subscriptions
How to code Claude Code in 200 lines of code
Opus 4.5 is not the normal AI agent experience that I have had thus far
Murder-suicide case shows OpenAI selectively hides data after users die
December 2025
2025: The Year in LLMs