sAimonSays.Ai
// module.models

one prompt does not fit all

Each model has its own dialect. Learn the quirks, the tag syntax, the weighting rules, and the tips that separate 30%-output from 90%-output.

OpenAI // GPT-5.x / GPT-4o

ChatGPT

go to ChatGPT
STRENGTHS
  • ▸Broad reasoning
  • ▸Tool use
  • ▸Code
  • ▸Structured output
QUIRKS
  • ▸May hedge without grounding
  • ▸Prefers explicit format instructions
BEST FOR
  • ▸Structured JSON
  • ▸Coding
  • ▸Multi-step agents
PRO TIPS
  • ▸Give a role + goal + constraints + output format in that order.
  • ▸Ask for 'citations or say I don't know' to reduce hallucinations.
  • ▸Use markdown headings for long-form outputs; it respects them well.
EXAMPLE PROMPT 3/4
You are a startup CTO reviewing a codebase you just inherited. Given this repo README + package.json, list: (a) 3 immediate technical risks, (b) 5 questions I should ask the previous team, (c) the first refactor I'd do. Be blunt.
quick nudge
Getting value from sAimonSays.Ai? Suggested donation is $20. Anything helps.
install as app
Add sAimonSays.Ai to your home screen for a full-screen, app-like experience.