The Question
ChatGPT now lets you set custom instructions — persistent preferences that shape how the model responds to you. Does that make a dedicated prompt tool like Alchemist redundant?
Short answer: no. They solve different problems.
What Custom Instructions Do
Custom instructions are persistent preferences. You tell ChatGPT things like "always respond concisely" or "I'm a software engineer, skip the basics." These shape the model's default behavior across all your conversations.
They're useful for setting baseline context. But they don't help you write better individual prompts.
What Alchemist Does
Alchemist works at the prompt level. When you're about to send a specific request, it helps you make that request clearer, more specific, and better structured — before you send it.
These are fundamentally different layers:
- Custom instructions: persistent, global behavior preferences
- Alchemist: per-prompt refinement and enhancement
Custom instructions might tell ChatGPT you're a developer. Alchemist helps you write a better request for what you actually need in this specific conversation.
The Combination Is Powerful
The best approach is to use both. Set your custom instructions to establish persistent context, then use Alchemist to polish and structure each specific prompt. You get the baseline behavior you want, plus the clarity and specificity that gets great outputs.
Coding Mode: No Equivalent
ChatGPT's custom instructions have no equivalent to Alchemist's Coding Mode — the feature that takes an app idea through a 5-step clarification flow and generates four ready-to-use prompt files. That's a qualitatively different capability.
Bottom Line
Custom instructions and Alchemist aren't competitors — they're complementary. Use custom instructions to set persistent preferences. Use Alchemist to write better individual prompts. Together, they significantly raise the quality of everything you get from AI.