Anthropic's change to Claude code's cache TTL from 1 hour to 5 minutes has led to increased quota usage and costs. This adjustment could impact developers relying on their API for cost management and performance optimization.
It looks like Anthropic changed claude code’s cache TTL from 1h to 5m in March, causing significant quota and cost inflation.
Google Cloud's GKE now supports native HPA for autoscaling without the need for adapters, reducing latency and costs. This change simplifies the scaling process, which could be relevant for engineers managing Kubernetes infrastructure.
Autoscaling on #GKE just got faster & cheaper!
Google removed the "middleman"—no more adapters or complex IAM for custom metrics.
Zero Adapters: Native HPA support
Lower Latency: Scale instantly
Cost Savings: No ingest fees
#GoogleCloud #Kubernetes #DevOps
This tweet outlines the official API pricing for several frontier AI models, including OpenAI's GPT-5.4 and Anthropic's Claude Opus 4.6. Senior engineers should care about these pricing structures as they directly impact cost management and decision-making for integrating these models into production systems.
Frontier models (Apr 2026 official API pricing, per 1M tokens):
- OpenAI GPT-5.4: $2.50 input / $15 output → $300 buys 120M input or 20M output tokens
- Anthropic Claude Opus 4.6: $5 input / $25 output → 60M input or 12M output
- Google Gemini 3.1 Pro: $2 input / $12 output