Curated weekly

Catch me up

What changed, and what it changes for you. Every item is triaged: act, verdict change, watch, or safe to ignore.

6 Important · 2 verdict changes · 0 FYI · 3 watching · 11 total

SImportantModels · Jun 22

Sakana Fugu: multi-agent orchestration as a model

Sakana AI's Fugu coordinates frontier models behind a single API. Fugu Ultra beats GPT 5.5 and Opus 4.8 on SWE Bench Pro at $5/$30 per 1M tokens. Verdict: Watch.

Read the full call

ImportantBreaking Changes · Jun 12

n8n 2.0 ships: community nodes break on upgrade, pin your version first

The new node API is worth the move, but self-hosted instances that auto-pull latest will lose community nodes silently. Pin, test, then migrate.

OImportantBreaking Changes · Jun 8

Using GPT-4 Turbo? It stops responding July 1. Migrate to GPT-4.1

OpenAI confirmed the shutdown date. Drop-in switch for most uses; function-calling payloads need one change.

ImportantPricing · Jun 22

Anthropic pauses Claude Agent SDK billing, for now

Anthropic halted plans to charge API rates for Claude Agent SDK usage on the day the change was to take effect. Subscription caps remain, but the pricing uncertainty is not going away.

ImportantIndustry · Jun 21

SpaceX Acquires Cursor for $60 Billion

Cursor is no longer independent. SpaceX acquires the AI editor for $60 billion. Here's what the ownership change means for developers who depend on it daily.

ImportantModels · Jun 21

Claude Fable 5: Launched, Then Banned by the US

Fable 5 launched as the top frontier model on June 9. Three days later, the US forced Anthropic to shut it down: the first-ever government ban of a deployed AI model.

Models

Foundation and frontier model releases, updates, and capability changes.

All Models news

GGVerdict changedJun 22

Z.ai ships GLM-5.2: open-source coding model, 1M context

GLM-5.2 scores 81.0 on Terminal-Bench 2.1 (+19 pts). IndexShare cuts per-token FLOPs 2.9x at 1M context. Open-source coding agents just got cheaper to run.

Read the full call

WatchingMay 28

Claude 4.6 Opus brings 1M context: long-document work just got simpler

If you chunk documents to fit context windows today, you can probably stop. Pricing is unchanged.

Product Launches

New product and tool launches entering the market.

All Product Launches news

VVerdict changedJun 22

Vercel Eve: open-source agent-as-directory framework ships

Vercel Eve ships production-grade agent infrastructure backed by 100+ internal agents, but every capability runs on Vercel. Free framework, paid runtime.

Read the full call

WatchingJun 12

Zapier Agents hits GA: wait for the pricing dust to settle

Agents that build and run Zaps from natural language are now generally available. Capable, but per-activity billing on top of per-task pricing needs a calculator.