Capy.ai launches a multi-agent cloud IDE — but does $20 get you anywhere?
Capy is a new cloud IDE that lets developers run up to 25 AI coding agents at the same time, each handling a separate task in its own isolated environment.
Capy is a new cloud IDE that lets developers run up to 25 AI coding agents at the same time, each handling a separate task in its own isolated environment.
Anthropic accidentally published the entire Claude Code source code to npm. A misconfigured build file did what no hacker could.
Alibaba just dropped another Qwen model on OpenRouter, shortly after Qwen 3.5 Omni. This one is text-only but packs a 1M token context window.
Hermes Agent’s biggest weakness compared to OpenClaw was single-agent limitation. v0.6.0 just fixed that.
Alibaba released Qwen3.5-Omni, an open-weight model that sees, hears, and speaks — and can turn a voice description into working code in real time.
Hermes Agent v0.5.0 is out. Nous Research calls it the hardening release — and it shows.
Ollama added a one-command way to launch a minimal coding agent. It is called Pi, and it powers OpenClaw under the hood.
Google DeepMind published research on AI manipulation. The study shows in which areas can AI manipulate you the most – and companies are already using these tactics to the fullest.
OpenClaw v2026.3.28 ships with 71 contributors and first-class xAI search, MiniMax image generation, and a new plugin approval system.
New research shows that in over 1 in 5 comparisons, the cheaper AI model actually costs more to run. The culprit is a cost that most providers do not clearly surface.