Gemini-SQL2: Google’s Text-to-SQL Clears 80% on BIRD
Google Research’s Gemini-SQL2 hits 80.04% execution accuracy on the BIRD benchmark — 7 points ahead of GPT-5.5. Here’s what it means for developers ...
Gemini Managed Agents: Deploy an AI Agent with One API Call
Google's Gemini Managed Agents launch a Linux sandbox with one API call. Here's how the ...
Claude Managed Agents: Cron Schedules and Credential Vaults, Explained
Claude Managed Agents now supports cron-based scheduling and credential vaults in public beta. Here is ...
Gemini 3.5 Pro: 2M Context, Deep Think, and the Flash-vs-Pro Decision
Gemini 3.5 Pro GA is days away. Here is the exact decision framework for when ...
Fable 5 Banned: US Gov Directive and Your Claude API Fix
Fable 5 and Mythos 5 are globally suspended after a US export control directive on ...
Kimi K2.7-Code: Moonshot’s Open-Weight 1T Coding Agent
Moonshot AI released Kimi K2.7-Code on June 12 — a 1T open-weight coding agent at $0.95/M tokens. Here’s how to use it in ...
Cursor Shared Canvases: From IDE to Team Dashboard
Cursor 3.7 added Design Mode to canvases. Combined with Shared Canvases, agent output now travels ...
Agentjacking: The Fake Bug Report That Hijacks Your AI Coding Agent
A new attack class lets anyone with your Sentry DSN hijack Claude Code, Cursor, or ...
Stripe MPP and x402: How to Bill AI Agents for Your API
Learn how to implement x402 and Stripe MPP to charge AI agents for your API ...
DeepSeek Retires deepseek-chat July 24: Migrate Now
DeepSeek retires deepseek-chat and deepseek-reasoner on July 24, 2026. Here is the one-line migration fix ...