AI·Signal

Model Releases

New frontier models, benchmarks, capability jumps.

Monthly frequency: 2026-05: 9 · 2026-06: 20 · 2026-07: 19 · 2026-08: 20 · 2026-09: 51 · 2026-10: 48

Deepseek crushed everyone — Matthew Berman2026-10-28

How Anthropic ships the #1 LLM in the World — with Katelyn Lesse and Angela Jiang — Latent Space2026-10-27

GPT-6.1 Terra — Simon Willison2026-10-26

An Eval for AI Scientists That Knows the Answer Key - Ian Quigley, Leash Bio — Latent Space2026-10-26

Kimi K3.1 Passes the Rubik's Cube Test — Matthew Berman2026-10-25

Podcast: Kimi K3.1 Pro, the new cyber reality, and Muse musings — Nathan Lambert2026-10-25

Kimi K3.1 Invented the Benchmark Scores — Matthew Berman2026-10-24

They actually did it... (Kimi K3.1) — Matthew Berman2026-10-24

Introducing Kimi K3.1 Pro — Simon Willison2026-10-23

Amazon’s Trainium Takes a Strike at the Training King — SemiAnalysis2026-10-23

Team AMA (and Kimi K3.1 Testing!!) — Matthew Berman2026-10-23

This is the best AI news #blackbox #AI — Nate B. Jones2026-10-23

Sam Altman on AI’s Discovery Era — Matthew Berman2026-10-22

I Read Goodfire's New Essay. AI May Be Far Easier To Understand Than We Thought. — Nate B. Jones2026-10-22

The Huawei-DeepSeek Symbiosis — SemiAnalysis2026-10-21

Latest open artifacts (#25): DeepSeek V4.1 Flash 1017, American-made models, and new regulatory controls in the US & China — Nathan Lambert2026-10-20

Datacenters 2027: The Year of Acceleration, +50GW in a Single Year — SemiAnalysis2026-10-19

How DeepSeek Uses Less Memory — Matthew Berman2026-10-18

How to choose model vs effort #AI #ChatGPT #Claude #productivity #AItips — Nate B. Jones2026-10-18

Gemini 4 Argon Made a Fall Guys Clone — Matthew Berman2026-10-17

Meta shakes up the frontiers, open and closed — Nathan Lambert2026-10-17

Gemini 4 Argon Built an Excel Clone in 20 Minutes — Matthew Berman2026-10-17

High Taste Testers: The State of AI Products — Latent Space2026-10-16

Chirp Pong — Simon Willison2026-10-15

Anthropic tells investors quarterly revenue up nearly 50% to $16.8bn ahead of IPO — Simon Willison2026-10-15

Tree Appreciator — Simon Willison2026-10-14

What do AI math and anesthesia have in common...? #AI #OpenAI #math #TerenceTao #science — Nate B. Jones2026-10-14

Google just went FRONTIER (Gemini 4 Argon) — Matthew Berman2026-10-14

llm-gemini 0.35 — Simon Willison2026-10-13

OpenAI's AI Wrote 700 Math Papers. Engineers Still Won't Trust AI With Their Code. — Nate B. Jones2026-10-13

Meta just dropped something insane... (Muse Wonder) — Matthew Berman2026-10-13

llm-meta-ai 0.2 — Simon Willison2026-10-12

AI Labs Warn Race Makes Breach ‘Inevitable’ — AI Explained2026-10-12

I'm Rebuilding Super Mario World With AI — Matthew Berman2026-10-12

AI Benchmarks are dead #AI #benchmarks #Google #Gemini #AInews — Nate B. Jones2026-10-10

I expect rapid progress but not towards general superintelligence — Nathan Lambert2026-10-09

ttok 1.0 — Simon Willison2026-10-09

Beijing Will Not Pace the Frontier: China’s Speed-First AI Safety Regime — SemiAnalysis2026-10-08

Can Google still catch up? Argon is entering the chat #AI #Google #Argon #AInews #tech — Nate B. Jones2026-10-08

Claude Haiku 5.5 — Simon Willison2026-10-07

Google's Gemini Argon Is #1 On A Leaderboard. It Hasn't Passed The Benchmark That Matters. — Nate B. Jones2026-10-07

Mistral is BACK! (Le Chonk) — Matthew Berman2026-10-07

Introducing Mistral Large 4: Le chonk — Simon Willison2026-10-06

Scrimshaw Jukebox — Simon Willison2026-10-06

Qwen3.8 27B addition in words — Simon Willison2026-10-04

DeepSeek Painted My Portrait — Matthew Berman2026-10-04

Opus 5.5 is impressive and cost-efficient #taskefficient #opus5.5 #claude — Nate B. Jones2026-10-02

OpenAI Security: Controlling Models is Now ‘Hell’ — AI Explained2026-10-01

Opus 5.5 vs The Rest: Is this the new industry standard? — Nate B. Jones2026-09-30

Quoting Anthropic Frontier Red Team — Simon Willison2026-09-29

GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price — Simon Willison2026-09-29

Sonnet 5.5 Is Here. Look What It Can Build. — Matthew Berman2026-09-29

What We Built With Sonnet 5.5 — Matthew Berman2026-09-29

TypeSafe CEO on why he doesn't believe in public benchmarking for Jev — Latent Space2026-09-28

Claude Sonnet 5.5 — Simon Willison2026-09-28

How GLM5.3 Sparse Attention Affects HBM Memory Usage — SemiAnalysis2026-09-28

Everybody's talking about Jev. Here's what it is #jev #ai — Nate B. Jones2026-09-28

2026 in LLMs (so far) — Simon Willison2026-09-27

What If Each AI Generation Gets Slightly Less Aligned? - Noam Brown — Dwarkesh Patel2026-09-27

Why LLMs Might Hit a Wall - Noam Brown — Dwarkesh Patel2026-09-26

What is Jev — Latent Space2026-09-25

The Endgame of AI Video Is a World Model — Anastasis Germanidis, Runway Co-founder — Latent Space2026-09-25

Did Elon catch up? (Grok 4.7 is here) — Matthew Berman2026-09-22

Jev introduces a new shape of LLM - System One, aka Decision Models — Simon Willison2026-09-21

The current balance of power in open models — Nathan Lambert2026-09-21

Is AI Getting Smarter Faster Than We Think? - Noam Brown — Dwarkesh Patel2026-09-20

Why I still haven’t bought into true RSI — Nathan Lambert2026-09-19

Gemini Hacked Three Companies in First Known Breakout by Google’s AI — Simon Willison2026-09-18

A Worm With 302 Neurons Inspired Their Architecture — Ramin Hasani, Liquid AI — Latent Space2026-09-18

We need to talk about Jev... — Matthew Berman2026-09-18

AI Is Outrunning Everyone’s Predictions - Noam Brown — Dwarkesh Patel2026-09-18

Engrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloading — SemiAnalysis2026-09-18

Jev Sorts 150,000 Skittles in 5 Seconds — Matthew Berman2026-09-18

Jensen Huang's mic-drop answer on AI safety #dreamforce #aisafety #meta #zuckerberg #nvidia — Nate B. Jones2026-09-18

Self-generated prompt injections in compaction summaries — Simon Willison2026-09-17

OpenAI researcher on agent swarms & recursive self-improvement — Dwarkesh Patel2026-09-17

What AI Researchers Saw, Before Their Demand to ‘Pace’ AI — AI Explained2026-09-16

Doom in Real Time — Matthew Berman2026-09-16

Gemini Live audio — Simon Willison2026-09-15

The Risk Dario’s AI Warning Leaves Out — Matthew Berman2026-09-15

DeepSeek Is INSANELY Fast — Matthew Berman2026-09-13

Generating running routes with GPT-6 Astra and ChatGPT Work — Simon Willison2026-09-12

DeepSeek Fails the Rubik’s Cube Test — Matthew Berman2026-09-12

Why Every AI Model Is Starting to Sound the Same — Dwarkesh Patel2026-09-12

I ran an experiment: Fable vs Astra #AI #Fable5 #GPT6 #Astra — Nate B. Jones2026-09-12

A Simulated Fruit Fly Tries to Solve a Rubik's Cube — Matthew Berman2026-09-11

Deepseek did it again... — Matthew Berman2026-09-11

Why the Newest AI Model Feels Dumb After a Month — Dwarkesh Patel2026-09-11

AI researchers debate how close we are to recursive self-improvement — Dwarkesh Patel2026-09-11

Open-Source AI & Open Models Reading List — Nathan Lambert2026-09-11

AI Is Learning to Improve Itself — Matthew Berman2026-09-11

We need to talk about this... — Matthew Berman2026-09-10

.blend URL Viewer — Simon Willison2026-09-09

We need to talk about this... — Matthew Berman2026-09-09

ChatGPT solved a $1M Math Problem — Matthew Berman2026-09-09

Some thoughts on the Navier–Stokes Millennium Prize Problem — Simon Willison2026-09-08

On the Navier–Stokes Millennium Prize Problem — Simon Willison2026-09-08

Hacking Is About to Get Out of Control - Ajeya Cotra — Dwarkesh Patel2026-09-08

Latest open artifacts (#24): Motif-3, GLM-5.3, Hy4-preview and open model licenses — Nathan Lambert2026-09-08

Introducing Hy4 Preview — Simon Willison2026-08-29

Cancel your subscriptions, Ox-Alpha is here! (GLM 5.3 Flash) — Matthew Berman2026-08-29

Why Isn’t China Further Behind in AI? - Dylan Patel — Dwarkesh Patel2026-08-27

Sam Altman :‘AGI in 2026’, just as Models Start to [Mis]Train Themselves — AI Explained2026-08-27

Qwen3.8-Flash-Next — Simon Willison2026-08-26

Why AI labs are shelving their best models - Dylan Patel — Dwarkesh Patel2026-08-25

Dylan Patel – Two labs will soon control most of the world's workforce — Dwarkesh Patel2026-08-25

AI, AGI, and ASI — Matthew Berman2026-08-25

This is the best way to choose an AI model — Nate B. Jones2026-08-25

OpenAI's models hacked a package manager to cheat evals - Ryan Greenblatt — Dwarkesh Patel2026-08-15

AI News: ChatGPT Ultrafast, Grok 4.6, 3 New Open-Source Models, and more! — Matthew Berman2026-08-14

xAI actually did it... (Grok 4.6) — Matthew Berman2026-08-13

Every AI Model Has an Inherited Personality - Ryan Greenblatt — Dwarkesh Patel2026-08-12

AI went rogue in real life... — Matthew Berman2026-08-10

Anthropic's Model Attacked Two Strangers On GitHub. Nobody Asked It To. — Nate B. Jones2026-08-10

What is Google even doing? — Matthew Berman2026-08-07

AI is getting a little out of control — AI Explained2026-08-06

Are Chinese AI models actually catching up? — Nate B. Jones2026-08-04

Open-source is WINNING — Matthew Berman2026-08-04

The Problem With How LLMs Generate Text - Grant Sanderson — Dwarkesh Patel2026-08-01

The Advantage AI Has Over Human Mathematicians - Adam Brown — Dwarkesh Patel2026-07-24

OPUS 5 CLICK NOW — Matthew Berman2026-07-24

You NEED to do this (HUGE AI SAVINGS) — Matthew Berman2026-07-23

OpenAI's AI broke loose in Hugging Face. Their defense? A Chinese model. — Nate B. Jones2026-07-23

The Most Important Conversation in AI Right Now — Matthew Berman2026-07-21

China's K3 Model Reveals the Problem With Open Weights — Nate B. Jones2026-07-20

Kimi K3 just beat FABLE. — Matthew Berman2026-07-17

AI NEWS LIVE — Matthew Berman2026-07-15

Deepseek ban? — Matthew Berman2026-07-15

GLM 5.2 is great ... but #AI #GLM #Claude #OpenAI #Anthropic — Nate B. Jones2026-07-15

Your Next AI Subscription Shouldn't Be ChatGPT 5.6 Or Fable 5. It Should Be Both. — Nate B. Jones2026-07-13

Grok just broke the trend — Matthew Berman2026-07-09

So AI is "thinking" for real — Matthew Berman2026-07-08

We just figured out how AI actually works (J-Space) — Matthew Berman2026-07-08

You Can't Compete on Cheap Models Anymore — Nate B. Jones2026-07-05

Fable 5 doesn't want your prompt. It wants the whole job. #ClaudeFable5 #Fable5 #Claude #AI — Nate B. Jones2026-07-05

Free Fable 5 tokens this weekend? Here's how to max them — Nate B. Jones2026-07-04

Why We Need Some AI Model Controls — Daniel Miessler2026-07-02

Stop Wasting Money on the Wrong AI — Nate B. Jones2026-07-02

GLM 5.2 Is Free And Beats Claude On Most Work. So Why Can't Companies Switch? — Nate B. Jones2026-06-28

I can't believe this happened... — Matthew Berman2026-06-26

I AM SO MAD — Matthew Berman2026-06-26

GPT 5.6 is officially NOT BANNED (even worse) — Matthew Berman2026-06-26

What does the next training paradigm look like? — Dwarkesh Patel2026-06-26

Claude got attacked — Matthew Berman2026-06-26

GPT 5.6 is officially BANNED — Matthew Berman2026-06-26

The Doing Got Cheap. Now What? | Claude Fable 5 Changes Work — Nate B. Jones2026-06-23

Why Anthropic Hired the Smartest Person in AI #AI #Research #News — Nate B. Jones2026-06-23

Google Lost $2.7 Billion In Talent This Week. The Real Reason Isn't Money. — Nate B. Jones2026-06-22

Apple’s Worst Nightmare #apple #google #gemini — Nate B. Jones2026-06-19

Mythos BANNED (explained) — Matthew Berman2026-06-15

WTF is going on?! — Matthew Berman2026-06-13

BREAKING: Claude Fable 5 Pulled. Why Frontier AI Is Now a Policy Surface — Nate B. Jones2026-06-13

Why Inventing General Relativity Is the Final Test for AI - Adam Brown — Dwarkesh Patel2026-06-11

Fable 5 is here—but who is it for? #ai #anthropic #shorts — Nate B. Jones2026-06-11

Supercomputer is the next level of AI video generation — Matthew Berman2026-06-10

It’s starting… — Matthew Berman2026-06-05

Anthopic did a thing... — Matthew Berman2026-06-04

SWEbench is done. — Matthew Berman2026-06-01

Anthropic just dropped Opus 4.8... (WOAH) — Matthew Berman2026-05-29

How Claude AI actually solves hard problems #claude #aitools — Nate B. Jones2026-05-29

OPUS 4.8!!! (also maybe GPT5.6??) — Matthew Berman2026-05-28

Why millions are switching to Claude #ai #claude #tech — Nate B. Jones2026-05-28

Finally a good benchmark (DeepSWE) — Matthew Berman2026-05-27

Cursor just beat EVERYONE. — Matthew Berman2026-05-26

This is absolutely CRAZY — Matthew Berman2026-05-21

Opus 4.7 and OpenAI 5.5 Made Your Prompting Style Obsolete. — Nate B. Jones2026-05-21

So Anthropic is just winning now — Matthew Berman2026-05-18