@brucethemoose You bought a GLM coding plan and hit throttling when it gets busy. A pay-per-token setup avoids busy-period throttling differently: live balance, top up from $1, one OpenAI-compatible /v1 key for GLM/DeepSeek/Qwen/Kimi, card / Apple Pay / Google Pay / PayPal, no region restrictions. Worth comparing for your workflow. Try it at api.getapiomni.com