The challenge

Which LLM to choose?

• GPT5 itself has 6 versions (including the OSS versions)
  • lowest cost = $0.05 per million input tokens and $0.4 per million output tokens for gpt5-nano
  • lowest cost including oss versions = $0.04 per million input tokens for gpt-oss-20b
  • highest cost = $1.25 per million input tokens and $10 per million output tokens for gpt5 (
    • 25x the lowest priced LLM

How to decide on the schema?

• tradeoffs involved in using a complex schema

How to engineer the prompt to get the most out of your investment?