Skip to content
BotServBotServ
ChatGPT promptsChatGPT commandsPrompt engineeringAI promptsChatGPT codesPrompts listOpenAI

100 ChatGPT Prompts & Commands: Ultimate List

100 ChatGPT prompts across 10 categories: coding, marketing, SEO, video editing, business, data analysis, social media and more.

S

schutzgeist

6 min read
100 ChatGPT Prompts & Commands: Ultimate List

100 ChatGPT Prompts & Commands: The Ultimate List for Coding, Marketing, Video & Business

What this article covers

  • 100 tested ChatGPT prompts and command codes across 10 categories.
  • Slash-style short commands (/optimize, /template, /matrix, /table), usable as prompt prefixes.
  • Per category: concrete commands with example prompts.
  • How to customize the codes and combine them with system prompts.

Introduction

The most effective ChatGPT users don’t work with long-form sentences. They use command codes: brief, precise directives that tell the model exactly what to do. /optimize, /table, /matrix, /template - these prefix-style prompts are reproducible, faster to type, and deliver more consistent results than free-form writing.

This list contains 100 such codes across 10 categories, with prompt text you can use directly. They work with ChatGPT (GPT-4o, o3) and can be adapted for other models. See also the Claude version and for local models the Ollama prompt engineering guide.

How the codes work

A command is a prefix + your input. Example:

/optimize
[Your text here]

ChatGPT interprets the command as an instruction and the rest as input. You can combine them: /optimize as table or /matrix with 3 options.

Category 1: Coding & Development (10 commands)

CommandWhat it doesExample prompt
/optimizeOptimize code (performance, readability)“/optimize [Code]: focus on runtime”
/debugFind and explain errors”/debug [Code + error message]“
/refactorRestructure code without changing behavior”/refactor [Code] to Clean Code”
/testGenerate unit tests”/test [Function] with pytest”
/documentWrite docstrings/comments”/document [Code] in Google style”
/reviewCode review with findings”/review [Code]: security + performance”
/explainExplain code line by line”/explain [Code] for beginners”
/regexBuild and explain regex”/regex for email validation”
/sqlGenerate/optimize SQL query”/sql for top 10 customers by revenue”
/apiDesign API endpoint”/api REST endpoint for User CRUD in FastAPI”

Category 2: Marketing & Advertising (10 commands)

CommandWhat it doesExample prompt
/copyWrite ad copy”/copy for [Product], target audience: young parents”
/campaignCampaign concept”/campaign for launch of [Product], budget 5k”
/taglineGenerate slogan/tagline”/tagline for [Brand], 5 variations”
/audienceBuild target persona”/audience for [Product]“
/funnelThink through funnel stages”/funnel for [Product], awareness to sale”
/ab-testCreate A/B variations”/ab-test for subject line [Text]“
/brand-voiceDefine tonality”/brand-voice for [Brand]: 5 adjectives + example”
/adAd copy for platform”/ad for Meta Ads, [Product], hook + CTA”
/newsletterNewsletter draft”/newsletter about [Topic], 3 sections”
/launchSketch launch plan”/launch for [Product], 4-week plan”

Category 3: Video Editing & Content Production (10 commands)

CommandWhat it doesExample prompt
/scriptWrite video script”/script for YouTube video about [Topic], 10 min”
/storyboardPlan shots/scenes”/storyboard for 60s product video”
/hookFirst 5 seconds of video”/hook for video about [Topic], 3 variations”
/thumbnailThumbnail idea + text”/thumbnail for [Video title]“
/chaptersGenerate chapter markers”/chapters from this transcript: [Text]“
/brollB-roll ideas list”/broll for [Topic] video”
/shortsShort version for Shorts/Reels”/shorts from this script: [Text]“
/voiceoverVoiceover text with timing”/voiceover for 30s clip, calm tone”
/title-videoVideo title variations”/title-video for [Topic], CTR-optimized”
/descriptionVideo description + tags”/description for [Video], SEO-optimized”

Category 4: Writing & Content (10 commands)

CommandWhat it doesExample prompt
/rewriteRewrite text”/rewrite [Text], more professional”
/expandExpand text”/expand [Text], double the length”
/shortenShorten text”/shorten [Text] to 100 words”
/toneChange tone”/tone [Text] → casual instead of formal”
/proofreadProofread and style”/proofread [Text]“
/simplifySimplify complexity”/simplify [Text] for non-experts”
/headlineHeadline variations”/headline for [Article topic], 5 options”
/introWrite introduction”/intro for article about [Topic]“
/conclusionWrite conclusion/summary”/conclusion for [Article summary]“
/translateTranslate with style”/translate [Text] into English, natural”

Category 5: SEO (10 commands)

CommandWhat it doesExample prompt
/keywordsKeyword ideas + intent”/keywords for [Topic], 20 items with intent”
/metaMeta title + description”/meta for page about [Topic]“
/outlineSEO article outline”/outline for [Keyword], H2/H3 structure”
/schemaSchema.org markup”/schema for FAQ page about [Topic]“
/interne-linksLink suggestions”/interne-links for article about [Topic]“
/titleTitle tag variations”/title for [Keyword], CTR-optimized”
/snippetFeatured snippet text”/snippet for question [Question], 40 words”
/content-gapFind missing subtopics”/content-gap in this article: [Text]“
/eeatE-E-A-T improvements”/eeat for article about [Topic]“
/lokal-seoLocal SEO copy”/lokal-seo for [Company] in [City]“

Category 6: Business & Productivity (10 commands)

CommandWhat it doesExample prompt
/swotSWOT analysis”/swot for [Company/Idea]“
/pitchElevator pitch”/pitch for [Product], 60 seconds”
/meetingAgenda + notes”/meeting agenda for team retro”
/emailProfessional email”/email to [Person], purpose: [X], tone: friendly”
/reportStructure report”/report from this data: [Text]“
/agendaMeeting agenda”/agenda for 30-min meeting about [Topic]“
/todoPrioritize tasks”/todo from this mess: [List]“
/brainstormIdeas list”/brainstorm 20 ideas for [Problem]“
/decisionDecision matrix”/decision between [Option A/B/C]“
/summarySummarize”/summary of [Text], 3 bullet points”

Category 7: Data Analysis & Tables (10 commands)

CommandWhat it doesExample prompt
/analyzeAnalyze data/text”/analyze these numbers: [Data]“
/matrixComparison matrix”/matrix [Options] by criteria [X,Y,Z]“
/tableOutput as table”/table [Data] with columns [A,B,C]“
/chartChart idea + code”/chart for this data in Python/matplotlib”
/compareA vs. B comparison”/compare [A] vs [B] by [Criteria]“
/forecastTrend estimate”/forecast based on [Data]“
/segmentSegment data”/segment this list by [Criterion]“
/surveyBuild survey questions”/survey for customer satisfaction, 10 questions”
/insightExtract insights”/insight from this report: [Text]“
/csvStructure as CSV”/csv this data: [Text]“

Category 8: Social Media (10 Commands)

CommandWhat it doesExample prompt
/postGenerate platform post”/post for LinkedIn about [topic]“
/threadX/Twitter thread”/thread about [topic], 5 tweets”
/captionInstagram caption”/caption for photo of [description]“
/hashtagsHashtag set”/hashtags for [topic], 15 pieces”
/bioProfile bio”/bio for [person/brand], 150 characters”
/replyReply to comment”/reply to this comment: [text]“
/viralViral framework post”/viral post about [topic], Problem-Agitate-Solve”
/pollPoll post”/poll for [community] about [topic]“
/carouselLinkedIn carousel content”/carousel about [topic], 8 slides”
/content-planWeekly content plan”/content-plan for [channel], 1 week”

Category 9: Learning & Education (10 Commands)

CommandWhat it doesExample prompt
/explain-einfachExplain like you’re 5”/explain-einfach [concept]“
/quizGenerate quiz questions”/quiz on [topic], 10 questions with answers”
/flashcardsFlashcards”/flashcards for [topic], front/back”
/lessonMini-lesson”/lesson about [topic] for beginners”
/tutorSocratic tutor”/tutor teach me [topic] through questions”
/analogyFind analogy”/analogy for [complex concept]“
/examplePractical examples”/example for [concept], 3 examples”
/study-planStudy plan”/study-plan for [topic] in 4 weeks”
/eli5Explain like I’m 5”/eli5 [topic]“
/fehlerCommon mistakes on topic”/fehler when learning [topic]“

Category 10: Prompt Engineering & Meta (10 Commands)

CommandWhat it doesExample prompt
/templateBuild prompt template”/template for recurring task [X]“
/frameworkSet thinking framework”/framework analyze with Porter’s Five Forces”
/personaAdopt role”/persona handle as experienced [role]“
/roleplaySimulate scenario”/roleplay customer conversation, I’m the seller”
/chainStep by step”/chain think step by step about [problem]“
/iteriereImprove in rounds”/iteriere this text 3x better”
/critiqueCritique output”/critique your own answer”
/improve-promptImprove prompt”/improve-prompt my prompt: [text]“
/systemSet system behavior”/system always respond as [role]“
/formatSet output format”/format answer as markdown table”

How to combine the codes

/persona handle as senior marketing consultant.
/template create a reusable template.
/matrix compare the 3 variants in a table.
/table show the result as a table with columns [variant, pro, con, recommendation].

Commands stack, each one adds a rule. Pro tip: save frequent combinations as Custom Instructions or GPTs so they’re always active.

For local models

The codes work with Ollama and local models too. Instead of repeating them per message, put them in the system prompt. See System Prompts and Prompt Engineering.

Further reading

Key takeaways:

  • Slash codes (/optimize, /template, /matrix) are reusable prompt shortcuts.
  • 100 commands in 10 categories: coding, marketing, video, copy, SEO, business, data, social, learning, meta.
  • Codes are combinable; each adds a rule.
  • For local models: put the same codes in the system prompt.
  • Save frequent combinations as Custom Instructions.

FAQ

Are these official ChatGPT commands?

No, they’re conventions that ChatGPT reliably understands. You can define your own codes; the model interprets /optimize or /table as a work instruction. The slash is just a readable prefix.

Do they work in every model?

Yes: GPT-4o, o3, Claude, Gemini, and local models (Ollama) understand the directives. Quality varies by model; with smaller local models, prefer simpler codes.

Can I combine codes?

Yes: codes stack rules. /persona sets the role, /table the format, /optimize the quality. Order doesn’t matter; all apply at once.

Define my own codes?

Yes, tell the model once “From now on, /x means: [instruction]” and keep using the code. Or save frequent combinations in Custom Instructions or GPTs permanently.

What about local AI (Ollama)?

Same codes, but put the rules in the system prompt so they apply persistently. See the Ollama System Prompts article.

Sources and further reading

Back to Blog
Share:

Related Posts