These are the 35 prompts worth saving for Grok 4.3, xAI's flagship since April 2026 with a 1M-token context window, native video input, and built-in document generation. They are grouped by job: real-time X search, cited DeepSearch research, writing, coding, business, and Grok Imagine visuals.

Every prompt is written out in full and ready to copy. The text prompts follow the Role + Task + Context + Constraints + Output format shape; the image prompts follow subject + environment + lighting + style + camera. Swap anything in [BRACKETS] for your own details. New to Grok's live search? The real-time X search guide goes deeper, and the Grok prompt cheat sheet is a handy one-pager.

Advertisement

Real-time X & trends

This is Grok's edge over other models: it reads X live. Start with "Check X for..." or "What are people saying on X about..." to trigger real-time search instead of the December 2025 training cutoff.

1. Live sentiment on a breaking topic

What are people saying on X about [TOPIC] in the last 24 hours? Summarize the overall sentiment (positive / negative / mixed), the three most common arguments on each side, and any notable shift in the last few hours. Quote 4-5 representative posts and name the accounts. End with a one-line takeaway.

Why it works: The "last 24 hours" window and the request to name accounts force Grok into live X search and keep the summary auditable.

2. What experts are posting right now

Check X for what credible experts and practitioners in [FIELD] are posting this week about [TOPIC]. Prioritize accounts with real domain track records over viral hot takes. Give me 6 bullet points of the actual substance being discussed, each with the account handle, and flag anything that contradicts the mainstream narrative.

Best for: Cutting past engagement bait to find signal from people who actually know the field.

3. Trending debate map

Check X for the current debate around [TOPIC]. Map it as a table: Faction | Core claim | Strongest evidence they cite | Weakest point | Example account. Cover at least three factions. Then tell me which position has the most momentum on X today and why.

Why it works: A table constraint pushes Grok to structure conflicting live posts instead of blending them into mush.

4. Real-time competitor watch

What are people saying on X about [COMPETITOR BRAND] in the last 7 days? Break it into: new announcements, praise, complaints, and any PR issues brewing. Quote the most-engaged posts in each bucket with handles, and end with two things [MY BRAND] could act on this week.

Best for: A fast, free competitive pulse without a paid social-listening tool.

5. Event live-thread digest

Check X for live posts about [EVENT / CONFERENCE / LAUNCH] happening now. Give me a running digest: the biggest announcements, the reactions that are getting traction, and any surprises. Update in reverse-chronological order with timestamps and handles so I can catch up in two minutes.

Why it works: Timestamps plus reverse-chronological order turn Grok into a fast catch-up feed for anything unfolding right now.

6. Emerging niche trends

Check X for early, still-small trends in [YOUR NICHE] that are gaining traction but haven't gone mainstream yet. I want signals, not obvious news. List 5, and for each give: what it is, why it's picking up, an example post with the handle, and how a creator or founder in this space could ride it early.

Best for: Spotting a wave before the crowd. Pair with the real-time X search guide for more search phrasings.

DeepSearch research

DeepSearch pulls and cites live sources instead of answering from memory. Ask for it by name and demand links, so you can verify every claim. Turn on Think mode when the reasoning is heavy.

7. Cited market landscape

Use DeepSearch to map the current landscape of [MARKET / INDUSTRY] as of this month. Cover: the main players and their positioning, recent funding or M&A, pricing norms, and where the market is heading. Cite every source with a link, prefer primary and recent sources, and note the publication date next to each fact.

Why it works: "Cite every source with a link" and "note the publication date" make the report checkable and current rather than a confident guess.

8. Fact-check a viral claim

Use DeepSearch to fact-check this claim: "[CLAIM]". Find the original source, what the primary evidence actually says, and how credible outlets have covered it. Rate it True / Misleading / False / Unproven with a short reason. Cite every source with a link, and separate what's confirmed from what's still uncertain.

Best for: Deciding whether something you saw online is real before you repeat it.

9. Literature and source roundup

Use DeepSearch to find the most important recent sources on [RESEARCH QUESTION]: studies, reports, and expert analyses from the last two years. For each, give a two-sentence summary, the key finding, and a link. Cite every source with a link, then end with a synthesis of where the evidence agrees and where it conflicts.

Why it works: Grok's 1M-token context lets it hold many sources at once, so the closing synthesis actually reflects the full set.

10. Compare products with sources

Use DeepSearch to compare [PRODUCT A] and [PRODUCT B] for a [USE CASE] buyer. Build a table: price, key features, real user complaints, and best-fit user. Pull from current reviews and forums, cite every source with a link, and end with a clear recommendation for someone who cares most about [PRIORITY].

Best for: A buying decision grounded in current reviews, not last year's marketing copy.

11. Regulation and policy brief

Use DeepSearch to brief me on the current rules for [ACTIVITY / TOPIC] in [JURISDICTION]. Cover what's required, what's changed recently, and what's proposed but not yet in force. Cite every source with a link, prefer official government and primary sources, and add a plain-English "what this means for me" line for each point. This is not legal advice.

Why it works: Anchoring to official primary sources and current dates is exactly where DeepSearch beats a static model.

Writing & content

Grok writes with a looser, more direct voice than most models. Give it a Role, a clear Task, real Context, and an exact Output format, and it will match your tone closely.

12. Blog post from an outline

You are a writer for [AUDIENCE]. Write a [1,200]-word blog post titled "[TITLE]" from this outline: [PASTE OUTLINE]. Keep the tone practical and direct, lead each section with the takeaway, use short paragraphs and concrete examples, and avoid clichés and filler. Output in clean Markdown with H2 sections and a two-line intro.

Best for: Turning a rough outline into a publishable draft that doesn't sound like every other AI post.

13. Sharp X thread writer

Write a 7-post X thread about [TOPIC] for [AUDIENCE]. Post 1 is a hook that earns the scroll without clickbait. Each following post makes one concrete point with a specific example or number. Keep each under 280 characters, no hashtags, no emoji spam, and end with a line that invites replies. Number the posts.

Why it works: Grok is native to X, so it has a strong feel for what reads well there; the constraints keep it tight.

14. Cold email in your voice

You are writing a cold outreach email from me ([MY ROLE] at [MY COMPANY]) to [PROSPECT ROLE] at [PROSPECT COMPANY]. Goal: [GOAL]. Keep it under 90 words, open with a specific reason I'm reaching out to them (not a template), state one clear value point, and end with a low-friction ask. Give me three subject-line options.

Best for: Outreach that reads like a person wrote it, with subject lines you can A/B test.

15. Rewrite for clarity and tone

Rewrite the text below to be clearer and more direct without losing meaning. Cut filler, fix weak verbs, and keep my key terms. Target a [confident but plain] tone at a [general professional] reading level. Return the rewrite first, then a short bullet list of the main changes you made. Text: [PASTE TEXT]

Why it works: Asking for the change list separately keeps the rewrite clean while still showing its work.

16. Headline and hook variations

Generate 12 headline options for [PIECE / PRODUCT] aimed at [AUDIENCE]. Mix angles: benefit-led, curiosity, contrarian, number-led, and how-to. Keep each under 65 characters, avoid hype words, and label the angle of each. Then pick your top three and say why they'd perform best.

Best for: A batch of testable headlines with the reasoning attached.

17. Newsletter from raw notes

Turn these raw notes into a [500]-word newsletter issue for [AUDIENCE]: [PASTE NOTES]. Structure: a strong one-line opener, 3-4 short sections with subheads, and a closing line with one clear call to action. Keep my voice casual and specific, and suggest a subject line and preview text. Output in Markdown.

Why it works: Feeding Grok your actual notes keeps the facts yours; it only handles structure and polish.

Advertisement

Coding & data

Grok 4 has native tool use and strong reasoning; flip on Think mode for hard debugging. Always paste real code and state your language, versions, and the exact error. For a deeper set, see the Grok coding prompts.

18. Debug with Think mode

Use Think mode. I'm getting this error in [LANGUAGE / FRAMEWORK version]: [PASTE ERROR]. Here's the relevant code: [PASTE CODE]. Walk through the likely root cause step by step, give me the corrected code, and explain what was wrong in two sentences. If more than one cause is possible, rank them by likelihood.

Why it works: Think mode spends extra compute on the reasoning, which pays off on subtle bugs; ranking causes keeps it honest.

19. Explain unfamiliar code

Explain what this code does to a developer who's new to [LANGUAGE / LIBRARY]: [PASTE CODE]. Give a one-paragraph overview, then a line-by-line or block-by-block walkthrough of the tricky parts, and call out any bugs, edge cases, or performance concerns you notice. Keep it concrete.

Best for: Onboarding to an unfamiliar codebase or reviewing something a teammate wrote.

20. Write tests for a function

Write thorough unit tests for this function using [TEST FRAMEWORK]: [PASTE FUNCTION]. Cover the happy path, edge cases, invalid inputs, and any boundary conditions. Explain in one line what each test verifies, and note any behavior in the function that looks like a bug or is ambiguous enough to need a spec decision.

Why it works: Asking Grok to flag ambiguous behavior surfaces spec gaps while it writes coverage.

21. SQL from plain English

Write a [PostgreSQL] query for this request: [WHAT I WANT]. Here's the schema: [PASTE TABLES / COLUMNS]. Return the query, a one-line explanation of each join or filter, and a note on indexes that would keep it fast at scale. If my request is ambiguous, state the assumption you made.

Best for: Getting a correct, performance-aware query without hand-writing joins.

22. Refactor for readability

Refactor this code for readability and maintainability without changing behavior: [PASTE CODE]. Improve naming, reduce nesting, and extract logic where it helps. Return the refactored version, then a short bullet list of what you changed and why. Keep it idiomatic for [LANGUAGE] and don't add dependencies.

Why it works: The "without changing behavior" and "no new dependencies" constraints keep the refactor safe to merge.

23. Regex builder and explainer

Build a regex for [LANGUAGE / FLAVOR] that matches [WHAT TO MATCH] but not [WHAT TO EXCLUDE]. Give me the pattern, a plain-English breakdown of each part, and 5 test strings (3 that should match, 2 that shouldn't) with the expected result for each. Keep it as simple as the task allows.

Best for: A working, explained regex with test cases instead of a cryptic one-liner you can't maintain.

Business & productivity

Grok 4.3 can generate documents like spreadsheets and slides, and it can pull live X signals into strategy work. Give it your real context and it handles the structure.

24. Meeting notes to action items

Turn these meeting notes into a clean summary: [PASTE NOTES]. Output three sections: Decisions made, Action items (as a table with Owner, Task, Due date), and Open questions. Keep it tight, don't invent details that aren't in the notes, and flag anything that was left unresolved.

Why it works: The owner/task/due table turns a wall of notes into something the team can actually track.

25. One-page business plan

You are a pragmatic startup advisor. Draft a one-page plan for this idea: [DESCRIBE IDEA]. Cover: the problem, the target customer, the offer, how it makes money, the first three steps to validate it cheaply, and the biggest risk. Be blunt about weaknesses. Output as clear sections, no filler.

Best for: Pressure-testing an idea before you sink time into it. The "be blunt" line matters.

26. Competitive SWOT with live X data

Build a SWOT analysis for [MY BUSINESS] competing against [COMPETITOR]. For the opportunities and threats, check X for what customers are currently saying about both of us and fold in real sentiment. Output a four-quadrant table, keep each point specific, and end with the single move I should make first.

Why it works: Blending a classic framework with live X sentiment gives you a SWOT grounded in what customers say right now.

27. Spreadsheet formula and model

I need a [Google Sheets] formula/model to [GOAL]. My data is laid out like this: [DESCRIBE COLUMNS]. Give me the formula, explain what each part does, and note common ways it could break (blank cells, text vs numbers). If a small helper column would make it cleaner, suggest it.

Best for: Getting an explained formula plus the gotchas, so you can maintain the sheet yourself.

28. Job description and hiring rubric

Write a job description for a [ROLE] at [COMPANY / STAGE]. Include: a two-line pitch for why the role matters, core responsibilities, must-have vs nice-to-have skills, and what success looks like at 90 days. Then give me a five-point scoring rubric to evaluate candidates fairly. Keep the tone human, not corporate.

Why it works: Pairing the JD with a rubric makes hiring more consistent and less gut-feel.

29. Prioritized weekly plan

Here's everything on my plate this week: [PASTE LIST]. Act as a sharp chief of staff. Sort it by impact and urgency, tell me the three things that actually move the needle, what to delegate or drop, and lay it out as a simple day-by-day plan. Call out anything that looks like busywork.

Best for: Turning an overwhelming to-do dump into a focused, honest weekly plan.

Grok Imagine (image & video)

Grok Imagine uses Aurora for stills and generates roughly 10-second 720p clips with audio; Agent Mode (beta) stitches 6-second clips into longer films. Lead with the subject, then environment, lighting, style, and camera, in 30-80 words. More recipes live in the Grok Imagine prompts guide.

30. Cinematic portrait still

A [weathered fisherman in his sixties] with a short grey beard, sitting on a wooden dock at [a quiet northern harbor] at dawn. Soft golden backlight and cool blue shadows, faint mist over the water. Photorealistic, warm cinematic color grade, shallow depth of field, shot on an 85mm lens at f/1.8, eye-level, tight head-and-shoulders framing.

Why it works: Subject first, one coherent photoreal style, and real lens details give Aurora a clear, single vision instead of adjective soup.

31. Product hero shot

A [matte black ceramic coffee mug] centered on a [smooth concrete surface] with a single sprig of eucalyptus beside it. Bright soft studio lighting from the upper left, gentle reflection below, clean neutral background with subtle shadow. Minimalist commercial product photography, crisp focus edge to edge, straight-on eye-level, square framing with generous negative space.

Best for: Clean e-commerce and marketing stills. Swap the product and surface to reuse it.

32. Stylized landscape

A [terraced rice valley in the mountains] at golden hour, thin clouds catching the last light, a small stone path winding through the fields. Warm amber and deep green palette, painterly semi-realistic style with soft brushed textures, gentle atmospheric haze in the distance. Wide establishing shot, slightly elevated vantage, deep focus across the whole scene.

Why it works: Naming one style ("painterly semi-realistic") and one palette stops Grok from blending clashing looks.

33. Character concept sheet

A [young space courier] in a worn orange flight jacket and patched cargo pants, holding a battered helmet under one arm. Neutral studio backdrop, even soft frontal lighting so details read clearly. Clean stylized game-concept-art look, full-body front view plus a smaller side view, sharp linework and flat coloring. Consistent design, eye-level, room around the figure for annotations.

Best for: A reusable character reference for games, comics, or storyboards.

34. Short animated clip

A [paper boat] drifting down a rain-slicked gutter after a storm, small ripples trailing behind it. Overcast afternoon light, cool desaturated palette with one warm reflection from a shop window. Gentle handheld camera slowly tracking alongside the boat, shallow depth of field, soft ambient rain and distant traffic audio. Calm, wistful mood, roughly ten seconds.

Why it works: Describing motion, camera movement, and audio explicitly is what turns a still prompt into a clean Grok Imagine clip.

35. Logo and brand mark concept

A simple modern logo mark for [BRAND NAME], a [what it does]. Combine [a mountain and a signal wave] into one clean geometric symbol. Flat vector style, two-color palette of [deep teal and warm sand], balanced and scalable, works at small sizes. Centered on a plain white background, even lighting, no text, no gradients, no clutter.

Best for: Fast logo directions to explore before commissioning a designer. Keep it flat and simple so it scales.

Want the shortest possible reference for all of this? The Grok prompt cheat sheet lists every mode and modifier on one page, and bookmark this roundup at /best-grok-prompts for the full set.

Frequently Asked Questions

Which Grok version are these prompts built for?

They target Grok 4.3, xAI's flagship as of April 2026, with a 1M-token context window, native video input, and document generation. Everything also runs on the Grok 4 family, since real-time X search, DeepSearch, and Think mode are shared features.

How do I make Grok search X in real time?

Start the prompt with "Check X for..." or "What are people saying on X about...". Those phrasings trigger Grok's live X search instead of answering from its December 2025 training cutoff. Add a date range and ask it to name the accounts it pulled from.

When should I use DeepSearch instead of a normal prompt?

Use DeepSearch when you need current, verifiable facts with sources. It pulls and cites live pages rather than answering from memory. Add "Use DeepSearch to... and cite every source with a link" so the output is auditable.

What is the text prompt formula for Grok?

Role + Task + Context + Constraints + Output format. Tell Grok who it is, what to do, the background it needs, the limits (length, tone, exclusions), and the exact shape of the answer, such as a table or a numbered list.

How do I write prompts for Grok Imagine?

Follow subject + environment + lighting + style + camera, in 30-80 words, leading with the subject. Aurora generates the still images and Grok Imagine turns prompts into roughly 10-second 720p clips with audio. Keep one coherent style and skip quality-adjective spam.

Can Grok generate video?

Yes. Grok Imagine produces about 10-second 720p video clips with audio, and Agent Mode (beta) batch-edits images and stitches 6-second clips into longer films. Describe motion and camera movement explicitly for better results.

Do I need the paid tier to use these prompts?

Most work on the standard tiers at grok.com or in the X app. Think mode and heavier DeepSearch runs benefit from a paid plan, and Grok 4 Heavy (about $300/mo) scales to 16 parallel agents for the hardest tasks. The prompts themselves are tier-agnostic.

How do the [BRACKET] placeholders work?

Anything in square brackets is a swap-in. Replace [TOPIC], [YOUR NICHE], or [CITY] with your own details before sending. The rest of each prompt is written to be copy-paste ready.

Advertisement