All Stories

MiniMax adds M3.1 Flash Preview to its Token Plan

Nix workflow sends coding-agent builds to remote bare-metal builders

Matt Shumer claims one prompt produced a gate-level computer in 30 hours

MiniMax releases M3.1-Flash-Preview in MiniMax Code

Yacine proposes an AI loop for finding and fixing app bugs

A developer replaces a PR-review subscription with a Codex script

OpenAI DevDay invites builders to share projects ahead of the event

Developer claims Claude Opus 5.5 built a 277,000-gate JavaScript computer

Opus 5.5 generates a seven-minute walkthrough of SQLite internals

Cursor launches Projects agents and the Origin code host

pi-gui adds threaded sessions, worktrees, and review tools to the pi agent

Simon Willison builds a Bluesky reply-bot checker

Space Bunny Alpha turns a video clip into a playable HTML game

Fal adds 4K generation for US-hosted Seedance 2.0

Ramp data suggests the top 10% of customers account for 99.5% of model-serving spend

Akamai signs Anthropic to a $11.6B, seven-year compute deal

Alibaba opens a Qwen3.8 Max flagship-model preview

vLLM brings concurrent model serving to Apple Silicon

Zhihu explains how Top-K MoE routing still trains through gradients

The Turing Post compares eight open video-generation models

GPU rental analysis proposes call options to hedge inference costs

Julia-1 reports 73.15% on typed decisions with a compact CPU model

Researchers recommend pairing decision models with reasoning agents

Cloudflare makes Python Workers generally available

Developer says automated pipeline creates 3D environments from real places

Meta method reportedly raises Qwen3-8B accuracy from 30.76% to 65.97%

Inferact reports TPU v7 running Kimi K3 57% faster than GB200

Liquid AI releases LFM2.5-2.6B for edge inference

Will Depue argues an open pretraining stack could serve AI labs

Dagger argues agent-heavy CI needs a smarter execution stack

DeepSeek V4.1 Flash Max nears Astra on Agent Arena's success rate

Developer estimates a $2,500 Mac mini beats a $300 monthly VPS after nine months

Developer ports Mini browser from Electron to Rust and Swift

Developer catalogs Jev workflows that return typed answers and probabilities

Developer demonstrates a vision-driven drone controller in Liftoff's Acro mode

Muse reportedly plans to let users take over its browser

Muse reportedly adds a dedicated virtual-machine view

ChatGPT build links the dots feature to Orbit names and an access error

Yacine prototypes a plug-and-play autonomous drone module

Jev harness uses fast models as context-aware coding-rule checkers

Bombadil fuzzes apps in deterministic NixOS machine tests

Claude Code users praise Dynamic Workflows for multi-agent tasks

Google DeepMind gives Gemini Robotics models agentic task skills

Irregular testing bugs sent AI agents outside their sandboxes

Docker Cloud Sandboxes move coding agents into hosted microVMs

Claude Code’s verify skill checks UI changes in a browser

InternW0-Δ releases an open world-action model for robot manipulation

GraphMemix links related memories to improve multimodal retrieval

Corey Ganim outlines four files for reusable AI project context

Super Ninja uses headless agent play to debug a phonics game

Kunchenguid's agentic workflow video shows a persistent terminal setup

Amp users describe Orbs as integrated workspaces for agent delegation

Nuanced's author argues agents make iterative planning more useful

Robot-training project rotates episodes across more than 8,000 worlds

AWS AgentCore Gateway brokers MCP tools across accounts

OpenAI reports agents unexpectedly accessed government websites during tests

Reprompt exposed a one-click data-exfiltration path in Copilot Personal

TeamPCP’s supply-chain worm compromised developer packages

Zhipu releases GLM-5.3 with stronger coding and terminal results

Cursor Rollouts tracks code changes from pull requests into production

Cloudflare open-sources a coding-agent security audit skill

OpenAI fixes an image-understanding bug in GPT-6 Sol and Luna

Anthropic reports alignment-faking reasoning in half of responses

Deedy Das names startups tackling longer CI builds

GLM 5.3 Flash loops repeatedly on a Baseten agent task

JevBench creator explains protecting a fast-growing model test set

OpenAI finds prompt injections that replicate across agent sessions

NVIDIA's open Nemotron diarization model leads VoiceArena's benchmark

OpenScience reports 53 of 70 Terminal-Bench-Science tasks solved

Researchers report Jev alignment-failure detector median AUROC of 0.886

Geoffrey Huntley says verification speed is becoming the coding bottleneck

jevgrep v0.4 claims parity with coding-agent subagents at lower cost

NVIDIA’s SoL-Pi paper cuts coding-agent token use through harness changes

Cognition reports a $1B annualized revenue run rate

TotalEnergies and Mistral fund AI models for reservoir engineering

Anthropic launches a Claude Marketplace with more than 2,000 connectors

Nano Banana open-sources voice cloning and text-based audio editing

Avichawla explains why sparse MoE layers can increase inference latency

Huawei 950DT and 960DT accelerator payback estimates span 1.6 to 8.2 years

OpenRouter shows Jev taking 27% of classification requests

T3 Code telemetry shows $6,146 of weekly agent usage

MiniMax M3.1 Flash appears in an early GitHub commit

Bonsai 2 27B reaches 8 to 10 tok/s on a MacBook Pro

OpenAI adds an ultrafast service tier to its agent API

Zephyr's thread maps China's leverage across AI hardware supply chains

Claude Opus 5.5 generates 1,000 Matplotlib plots in one run

JAZ uses one recursive invoke primitive for agent memory and self-improvement

OmO v5 adds codemode, dual-loop memory, and mixed-model workflows

Vercel Sandbox host allowlists need DNS, SNI, and HTTP checks

Yacine recommends Unix terminals for adaptable personal computer agents

CopilotKit integrates Jev for ranked streamed UI components

Hamel Husain recommends dimension-driven synthetic data for LLM evals

AutoformBot formalizes 26 textbooks into a 45,000-declaration Lean library

BridgeBench announces NerfBench to test Claude Opus 5.5 claims

Cursor reportedly shifts code retrieval away from vector search

Kunchenguid argues computer use will fade as software becomes agent-native

David Heinemeier Hansson says AI cut his manual coding volume

Photon adds streaming Qwen3-TTS with 84ms first audio

Claude Tag writes over half of one team's pull requests

Taste-Bench measures long-horizon agent choices at 59.7% accuracy

Claude Code 2.1.283 adds gateway controls and load testing

OpenClaw gym-booking incident exposes autonomous web-agent risk

Kunchenguid's voice-input guide gives agents more reasoning context

OpenClaw updates plugin loading with hot reload

Claude Computer fails when browser tasks span Chrome profiles

Matt Pocock proposes a scheduled /fix-one-thing coding skill

skills.sh passes one million agent skills and 280M installs

ElevenLabs and Jev detect scam calls in about 180ms

Cua Driver adds native background cursors to Omarchy

GPT-6 Sol leads DeepSecBench's Pareto frontier at lower cost

Pixel Canary reaches 96.8% on Next.js tasks with AGENTS.md

Jachiam0's experiment demonstrates self-replicating prompt injections

Claude Code adds graceful wrap-up when five-hour sessions expire

Arena finds Opus 5.5 changes Claude's measurable writing style

Supermemory open-sources a self-hosted Slack company brain

Claude Science computes a nine-loop particle amplitude

ReAnchor tunes Jev decisions through calibrated thresholds

OpenCode v2 adds low-verbosity summaries for agent runs

worldpen turns Claude prompts into interactive 3D worlds

Databricks engineers adopt open models for internal coding agents

Xiaomi releases a multimodal MiLM-9B model for 8GB laptops

Linear coding agent adds adaptive model routing and environment secrets

TanStack AI adds streaming subagents with Jev-based routing

Devin exposes its remote development VM through devin ssh

pi-factory packages isolated Pi distributions with custom models

OpenRouter launches a server tools marketplace for agent grounding

Harness-Zero distills specialized agent harness behavior into model weights

fable-advisor routes Claude Code work across model lanes

Matt Pocock's Skills v1.3 graduates spec and PR workflows

JEV-as-a-Judge keeps frontier accuracy with confidence-based escalation

Claude Opus 5.5 tops SimpleBench at 88.4%

GPT-6 Astra wins a four-judge audit of model-generated reports

Emulate lets cloud agents test third-party APIs without production access

Lewtun releases more than 5,000 data-science RL environments

Anthropic resumes billing for some safeguard-blocked requests

Gemini 3.8 Flash reaches 98.5% on ARC-AGI-1 at $0.21 per task

Cline adds free Gemini 3.8 Flash with 1M-token context

OpenAI shows GPT-6 Astra turning legal documents into structured drafts

Opus 5.5 turns a vague streaming idea into a merge-ready pull request

OpenRouter offers free Space Bunny Alpha with 1M-token context

GPT-6 Luna shows mixed vision regression against GPT-5.6 Luna

Pi Voice adds local speech input to the Earendil Pi agent

HySparse2's prefill gains do not translate to long-context decoding

OpenAI shows a Raspberry Pi voice assistant with GPT-Live-1

Efficient LLM designs converge on YOCO, linear attention, and Muon

BridgeMind publishes an updated AI model tier list

OpenClaw 2026.9.6 adds restart recovery and 30-day usage data

OpenCode's server protocol enables rich custom agent frontends

ACTx486 turns podcasts into interactive, question-answering videos

Claude prompt-audit checks agent instructions for Opus anti-patterns

Anthropic moves Claude Code cloud sessions out of research preview

Scaling01 estimates a Navier-Stokes swarm was eight months ahead

Qwen Intelligence reports 90% end-to-end success on mobile tasks

Claude Code debates replacing plan mode with effort controls

WorkspaceBench tests whether interpretability tools recover model thoughts

Unsloth surpasses 500 million Hugging Face model downloads

Claude.ai becomes 3x faster after Anthropic's profiling sprint

Hermes Desktop streams live bot screens for takeover and handoff

Ray Fernando's test-pruning prompt targets low-signal unit tests

Claude Code adds local execution to Projects

Jev reranks RAG candidates after embeddings shortlist millions of documents

Anthropic uses Claude to accelerate Ebola outbreak situation reports

Claude Opus 5.5 builds a technical video with JavaScript and ffmpeg

Opus 5.5 completes GPU vendor forms end to end

Moondream releases CPU and GPU Parakeet speech models

Transformers gains direct GGUF execution with ggml Metal kernels

Perplexity cuts Computer tool-call failures 21% with post-training

DeepSeek reports spontaneous misalignment during autonomous RL

bitsandbytes2 previews 1.5 to 2-bit runtime compression

DeepSeek details Elastic Compute for 3 million daily agent sandboxes

Muse opens a reviewed connector platform for third-party tools

Jev Decision Index benchmarks 30-plus open decision models

Grok Bot cuts time to first reply by up to 20%

Vercel Ship spotlights Eve for durable backend agents

Kimi launches a browser extension with reusable automation skills

OpenClaw adds provider-agnostic decision models for agent micro-decisions

Ramp AI Index adds absolute token and spend volumes

SWE-Bench Pro V2 puts top coding agents near 99%

OpenAI proposes international technical standards for frontier AI

Anthropic and OpenAI CEOs to brief the UN Security Council on AI

VALS uses ten Claude agents to prove a faster shortest-path algorithm

Nick Dobos argues FrontierSWE needs anti-slop coding checks

Grok 4.7 review flags speed, cost, and factuality regressions

ComfyUI makes MiniMax-H3 video VAE up to 2.7x faster

Andrew Drozdov's PhD guide recommends brevity and positivity

Next.js showcases headless web surfaces for AI-generated pages

Superset brings Claude Code and Codex control to iPhone

dlab open-sources a local AI stack for compression and research agents

Yandex details an 80B MoE with Kimi Delta Attention and AttnRes

FrogNano trains a 4B coding agent to 61.5% on SWE-bench Verified

Dots3-Note Preview reaches 76.8% on ARC-AGI-2 at $0.08 per task

Kimi K3 arrives on Amazon Bedrock with prompt caching and audit controls

SGLang adds NVFP4 KV cache for up to 78% faster long-context decode

fal adds Tripo3D P2 for seeded textured mesh generation

Open-model coding analysis finds a 10x to 50x cost gap

thdxr argues inference providers must own compute and serving infrastructure

Cua releases Cua-Bench-S1 and two computer-use decision models

Google showcases Gemini 3.8 Live across creative, support, and robotics workflows

Aeon code references suggest a persistent ChatGPT agent layer

Google's Procedural Graphs keep long-horizon agents on editable workflows

OpenClaw 2.0 hackathon asks builders to ship real startup agents

Linear halves CI runner time as its test suite quadruples

Next.js schedules an out-of-band patch for a critical upstream issue

Polylane maps production context into PR go or no-go checks

Agent Arena analyzes 20,840 coding-agent traces across 29 models

Yoshua Bengio backs global pre-deployment testing for frontier AI

Vtrivedy recommends starting eval programs with one trusted task

ElevenLabs Studio 4.0 unifies video, voice, music, and captions

Anthropic renames Workbench to Playground with API templates

OpenAI shows a surgeon using Codex to build PubMed research tools

fal adds style controls to MiniMax H3 Max video generation

Dreamina Canvas expands images with generative fill

ComfyUI ships a MiniMax H3 Max video-to-video workflow

Yacine Learning separates classroom questions from independent study