COSTAR vs RISEN: Which framework should you use?
A side-by-side breakdown of two of the most popular prompt frameworks — and a simple rule to know which fits your task.
PromptIQ Pro analyses your prompt in real time against 8 expert frameworks. See exactly what's missing, fix it in one click, and get better outputs — every time. For professionals, students, and everyone learning to use AI well.
Good prompt · minor improvements possible
Stop wasting time on back-and-forth. Score your prompt before you send and get the result you need first time.
PromptIQ Pro teaches you what a strong AI prompt looks like — in real time, as you type. Build a skill that matters for life.
Show students exactly what separates a weak prompt from a strong one. No theory — just live, visible feedback they can act on.
Score updates as you type, debounced to 300ms. No button to click, no waiting. See your grade before you send.
One click rewrites your prompt with bracketed placeholders for every missing element. Fill in the blanks, send a stronger prompt.
Rewrites your prompt in Claude's XML style, ChatGPT's markdown structure, or Gemini's natural language — optimised for each model.
Automatically picks the best framework for your prompt type — coding, creative, research, or business tasks.
See which elements are present or missing with evidence highlighted directly from your own text.
All scoring runs locally in your browser. Your prompts are never stored, logged, or transmitted to any server.
Open Claude, ChatGPT, or Gemini. Start typing your prompt as usual. PromptIQ Pro activates automatically.
A score badge appears in real time. Click it to see your full breakdown — grade, framework, and element checklist.
Click Fix My Prompt to fill in what's missing. Or adapt it for your target platform. Then send a better prompt.
| Framework | Best for | Key elements |
|---|---|---|
| COSTAR | Creative & detailed prompts | Context · Objective · Style · Tone · Audience · Response |
| RTF | Quick, focused prompts | Role · Task · Format |
| RISEN | Coding & technical tasks | Role · Instructions · Steps · End goal · Narrowing |
| CRISPE | Research & analysis | Capacity · Insight · Statement · Personality · Experiment |
| APE | Explanations | Action · Purpose · Expectation |
| CARE | Business outcomes | Context · Action · Result · Example |
| TAG | One-liners | Task · Action · Goal |
| BROKE | Strategy & planning | Background · Role · Objectives · Key results · Evolve |
Enterprise
Custom seats, SSO, procurement support, SLAAll scoring runs locally using pattern matching — no API calls, no server, no account required to try. Safe for corporate and confidential use.
Practical guides, framework deep-dives, and real examples to help you get more from AI.
A side-by-side breakdown of two of the most popular prompt frameworks — and a simple rule to know which fits your task.
We ran 100 common prompts through PromptIQ Pro. Here's what separates the A-grades from the failures, and what you can add in 30 seconds.
A practical lesson plan using live prompt scoring to show students — visually — what makes a strong prompt vs. a weak one.