Haiku 5.5 and GPT-6.1 Terra: the month LLMs became a routing problem

October 2026 was the month the cheapest useful language model became the story. Anthropic's Claude Haiku 5.5 arrived on 7 October with a 1 million-token context at $0.10 per million input tokens, OpenAI's GPT-6.1 Terra landed on 26 October at $1 input and $5 output per million tokens, and tiny open decision models started doing the routing itself. The question for builders is no longer which model to use. It is which step of each workflow deserves which model — and how to prove the cheap one passes before you trust it.