prompt-caching

How to manage, compress, and strategically fill the context window in long-horizon agents — covering summarization checkpoints, scratchpad memory, retrieval injection, prompt caching, and compaction triggers.

How frontier LLM costs and latency actually work in 2026 — input vs output token asymmetry, prompt caching, batch API discounts, TTFT and tokens-per-second, reasoning-model amplification, and when self-hosting breaks even.