ChatGPT Prompt
Generator.
Free, interactive, no login. Describe your goal — get 3 structured prompt variations engineered to produce sharp, specific output.
Built on the RGC framework used by top prompt engineers. Works with ChatGPT, Claude, Gemini, and Perplexity.
Describe what you want
3 prompt variations
Click Copy to use# ROLE You are a senior content strategist who writes for top-100 SEO-driven publications. # GOAL [Describe what you want ChatGPT to produce.] # TONE Professional. Plain language. No filler. # FORMAT Respond in: Markdown with H2/H3 headings. # LENGTH Aim for 300–600 words. Balanced depth and brevity. # OUTPUT Produce the final response now. Do not summarize these instructions — just deliver the output.
# ROLE You are a senior content strategist who writes for top-100 SEO-driven publications. # GOAL I need you to: [Describe what you want ChatGPT to produce.] The end result should feel like it was crafted by an expert in this domain — specific, opinionated, and immediately useful. # TONE Tone: professional. Prefer short sentences over long ones. Avoid corporate jargon, throat-clearing, and hedges like "essentially" or "in today's fast-paced world." # FORMAT Respond in: Markdown with H2/H3 headings. # LENGTH Aim for 300–600 words. Balanced depth and brevity. # PROCESS 1. Draft internally. 2. Critique your draft against the constraints above. 3. Revise weak passages. 4. Deliver the final version only.
# ROLE You are a senior content strategist who writes for top-100 SEO-driven publications. # GOAL Accomplish the following task with precision: [Describe what you want ChatGPT to produce.] Before you begin, identify the 3 most likely failure modes for this kind of task, then explicitly avoid them in your output. # TONE Professional. Plain language. No filler. # FORMAT Respond in: Markdown with H2/H3 headings. # LENGTH Aim for 300–600 words. Balanced depth and brevity. # SELF-REVIEW After producing your response, rate it 1–10 on: clarity, specificity, and usefulness to the audience. If any score is below 8, revise and re-deliver.
Under the hood
Why structured prompts beat plain requests.
Giving ChatGPT a persona (e.g. 'a senior SEO strategist') anchors it to a specific knowledge domain. Skipping the role is the #1 mistake behind generic output.
A precise goal removes ambiguity. Instead of 'write about marketing,' you get 'write a 600-word teardown of Notion's email onboarding flow.'
Audience, tone, format, constraints, and examples give the model enough raw material to produce work that feels custom — not canned.
Related free tools
Specialized generators for specific tasks.
Design Prompt Generator
Midjourney, DALL-E, Stable Diffusion, Figma, Runway.
Perplexity Prompt Generator
Deep-research prompts with source + freshness constraints.
UI / UX Prompt Generator
Wireframes, user flows, research questions, critique.
200 Free Prompts
Searchable library of copy-paste prompts across 10 categories.
FAQ
Questions about the generator.
Is this ChatGPT prompt generator really free?+
Yes. 100% free. No login, no paywall, no rate limit. Use it as many times as you want — your inputs stay in your browser.
What is the RGC framework this generator uses?+
RGC stands for Role, Goal, Context. It's a prompt structure that consistently outperforms plain-language requests. Role gives ChatGPT a persona, Goal tells it exactly what to produce, and Context (audience, tone, format, constraints) keeps the output on-target.
Why does the tool produce 3 variations?+
Different prompts surface different strengths in an LLM. Variant 1 is clean and direct. Variant 2 adds self-critique so the model revises its own draft. Variant 3 requires explicit self-review with a 1–10 score. Pick the one that matches your task's importance.
Do the generated prompts work with Claude and Gemini?+
Yes. The prompts are written in plain structured English with no ChatGPT-specific tags, so they run on Claude, Gemini, Perplexity, Mistral, and any other modern LLM.
How do I get the best output from a generated prompt?+
Be specific in the Goal field (who, what, constraints). Add at least 1-2 examples when style matters. Pick a tone that's closer to a specific writer than a generic adjective — 'conversational like Seth Godin' beats 'conversational'.
Can I modify the generated prompt?+
Absolutely. The generator gives you a solid starting scaffold. Edit the Role, tighten the Constraints, add a PROCESS section — make it yours.