ddd
17 posts
GLM has a serious token leakage / caching-accounting issue on Z.ai I tested this across Claude Code, Hermes Agent, Zcode, and OpenCode, so it does not look like one harness behaving badly. This has been consistent since the start of my subscription, during normal off-peak usage. My read: repeated context is being billed as fresh input instead of cached input. That’s not max reasoning. That’s a server-side caching/accounting problem. The screenshots show the issue clearly: Zcode used around 270K tokens, but I was billed for nearly 5M tokens. Cached tokens are clearly not working. I contacted support, escalated this, and tried everything, but got no real response. This is not how you treat paying customers. Please fix this. @ZixuanLi_ @Zai_org
More of the iOS app loop, now inside Codex. The Build iOS Apps plugin lets Codex view and test your iOS app in the in-app browser, open SwiftUI previews, and hot reload edits without leaving Codex.
🦊 Mole 1.5.0 is live, mole.fit A native Mac cleaner that now feels more like a small system companion. New in this release: • Menu bar monitor: CPU, memory, network speed at a glance, with a tiny runner that moves with your Mac's state • Fan control: Auto / Cool / Quiet modes with live RPM • App updates: check and install via Homebrew, App Store, and Sparkle • Startup manager: Login Items, Launch Agents, Daemons in one place • Smarter uninstall: alias search, input method cleanup $9 once, lifetime updates. MOLELOVE 20% off, ends tomorrow
Building an iPhone app directly in Codex desktop with iOS simulator
🚨 Did Moonshot just crack Continual Learning? Moonshot introduced: 'Checkpoint Engine'. It rapidly updates model weights in LLM inference engines. Can update a 1T param model like Kimi-K2 in 20 seconds. If this works, it truly is a *game changer* BIG. IF. TRUE. 🤯
Kimi K2 is number one trending on HF, congrats!
