Article

Jev means structured output is interesting again

Sean Goedecke · seangoedecke.com

A clear, sceptical explainer of decision models: why answering with a choice instead of text is fast and cheap, and where ordinary structured output already does the job.

Open on seangoedecke.com ↗

It answers 2 questions

  1. How do founders ship their first AI feature (APIs, wrappers, RAG in plain terms)?
  2. How much do AI agents cost to run, and how do I keep the bill under control?

Read with it

  1. OpenAI's endpoint for probability, choice and score answers on GPT-6 Luna, about 10x faster than the Responses API, for routing... developers.openai.com Guide Decisions API guide OpenAI
  2. The single chart that explains why what is unaffordable today is cheap next year. a16z.com Article Welcome to LLMflation: LLM inference cost is going down fast a16z (Guido Appenzeller)
  3. The headline numbers in thread form for a two-minute read. x.com Thread LLMflation: 10x cheaper per year a16z
  4. Names the paradox every founder hits: cheaper tokens, bigger bills. navyaai.com Article AI Token Cost Over Time: Down 99.7%, Bills Up 3x NavyaAI
  5. Explains the quadratic context growth that quietly multiplies agent costs. leanopstech.com Article AI Agents Burn 50x More Tokens Than Chats LeanOps
  6. How large buyers model agent economics, useful when selling to them. ey.com Article Agentic AI Enterprise Token Cost EY
eChai Partner Brands The partnership →