OnboardingRe: Welcome to DeepInfra! Start by changing two linesExplains how prompt caching affects latency and input-token costs, then shares ways to improve cache hits and inspect request usage. It also points to the full guide and mentions discounted Batch API processing.