cost-benchmark

Event-driven LLM request pipeline with a Postgres queue, idempotent ingress, retry/DLQ/replay — and the paired benchmark showing Haiku 4.5 cost 2-7x more per task than Sonnet 5 despite half the per-token price, which is why there is no cost router in src/.

Coding agent pricing comparison: cheapest LLM API for coding, cost per million tokens, from the OpenRouter price list as JSON and CSV. An AI model token price comparison repriced at a real agent mix — 95.6% cached input pricing, 6.5x off list: prompt caching price, cache read price, cache hit share, long context pricing, DeepSeek peak/off-peak.