These are the 40 prompts worth saving for DeepSeek-V3.2, the free open-source model released in December 2025 with a 128K-token context window and top-tier reasoning, math, and coding. They are grouped by job: reasoning, coding, math, writing, research, business, study, and everyday tasks. Run them at chat.deepseek.com or in the iOS and Android apps.

Every prompt is written out in full and ready to copy. The text prompts follow the Role + Task + Context + Constraints + Output format shape, but the golden rule with DeepSeek is different from most models: give it a hard problem plus rich context, not rigid formatting rules, and let it reason. Say "think step by step," paste real code, data, or source text, and turn on DeepThink for the hard ones. New here? The reasoning & math guide goes deeper, and the DeepSeek prompt cheat sheet is a handy one-pager. Swap anything in [BRACKETS] for your own details.

Advertisement

Reasoning & problem-solving

This is DeepSeek's edge. Turn on DeepThink and it shows its full chain-of-thought before the answer, so you can audit the logic. Feed it the hard problem and rich context, then ask it to verify against the constraints before committing.

1. Reason through a hard decision

Turn on DeepThink. I need to decide between [OPTION A] and [OPTION B]. Context: [YOUR SITUATION, CONSTRAINTS, WHAT MATTERS MOST]. Think step by step: lay out the real trade-offs, the second-order consequences of each, and the assumptions I might be wrong about. Then give a recommendation with your confidence level, and state what new information would change your answer.

Why it works: DeepThink exposes the reasoning so you can check where it agrees or disagrees with your gut instead of getting a black-box verdict.

2. Solve a logic puzzle step by step

Turn on DeepThink. Solve this logic problem and show every step of your reasoning: [PASTE PUZZLE]. Track the constraints as you go, note each deduction and why it follows, and eliminate options explicitly. Before giving the final answer, verify it satisfies every constraint in the problem. If more than one solution fits, say so.

Best for: Puzzles, riddles, and constraint problems where a wrong intermediate step ruins the answer.

3. Root-cause a messy problem

Turn on DeepThink. Here is a problem I can't pin down: [DESCRIBE SYMPTOMS, WHAT YOU'VE TRIED, RELEVANT DATA OR LOGS]. Reason step by step from symptoms to likely root causes. Rank the causes by probability, explain the evidence for each, and give me the single cheapest test to confirm or rule out the top one. Separate what you know from what you're inferring.

Why it works: Ranking causes by probability and naming the cheapest test turns vague reasoning into a concrete next action.

4. Stress-test a plan for flaws

Turn on DeepThink. Here is my plan: [PASTE PLAN]. Act as a sharp skeptic. Think step by step and find the ways this fails: hidden assumptions, weak links, single points of failure, and second-order effects I haven't considered. For each risk, rate its likelihood and impact, and suggest a concrete mitigation. Be blunt; I'd rather hear it now.

Best for: Pre-mortems on a launch, a project, or a big commitment before you're locked in.

5. Weigh options with a scored matrix

I'm choosing between these options: [LIST OPTIONS]. My criteria, most important first: [LIST CRITERIA]. Build a weighted decision matrix: score each option 1-5 on each criterion, apply the weights, and total them. Show the table, explain any score that isn't obvious, and give the winner. Then tell me which single criterion, if I'm wrong about its weight, would flip the result.

Why it works: Forcing weights and a sensitivity check makes the decision explicit instead of a rationalized gut call.

Coding

DeepSeek-V3.2 is top-tier at coding. Paste real code, state your language and versions, and give the exact error. Flip on DeepThink for gnarly bugs. For a deeper set, see the DeepSeek coding prompts.

6. Debug with a reasoning trace

Turn on DeepThink. I'm getting this error in [LANGUAGE / FRAMEWORK version]: [PASTE ERROR]. Here's the relevant code: [PASTE CODE]. Reason step by step through the likely root cause, give me the corrected code, and explain what was wrong in two sentences. If more than one cause is possible, rank them by likelihood. Verify your fix against the error before giving the final answer.

Why it works: DeepThink spends real compute on the trace, which pays off on subtle bugs; the verify step catches confident-but-wrong fixes.

7. Explain unfamiliar code

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

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

8. 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 boundary conditions. Explain in one line what each test verifies, and flag any behavior in the function that looks like a bug or is ambiguous enough to need a spec decision.

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

9. Refactor without changing behavior

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.

Best for: Cleaning up code you have to keep working in, safely enough to merge.

10. Build a feature from a spec

Turn on DeepThink. Build [FEATURE] in [LANGUAGE / STACK]. Requirements: [LIST REQUIREMENTS]. Constraints: [PERFORMANCE, DEPENDENCIES, STYLE]. First outline your approach and the key decisions, then write the full, runnable code with comments on the non-obvious parts. List any assumptions you made and how I'd test it. Verify the code meets each requirement before finishing.

Why it works: Outlining the approach first, then coding, catches design mistakes before they're buried in a wall of code.

Math & logic

Math is where DeepThink earns its keep. Turn it on so the full working is visible, and always ask DeepSeek to verify the answer against the constraints before it commits. A visible trace is easy to spot-check.

11. Solve a math problem with verification

Turn on DeepThink. Solve this problem and show all working: [PASTE PROBLEM]. State any formulas or theorems you use and why they apply. After reaching an answer, verify it by a different method or by substituting back into the original problem. If the check fails, find the error and redo it. Give the final answer clearly at the end.

Why it works: A second-method check catches arithmetic slips that a single pass would ship as a confident final answer.

12. Prove or disprove a statement

Turn on DeepThink. Prove or disprove: [STATEMENT]. Reason step by step. State your assumptions and definitions first, then build the argument rigorously. If it's false, give a clear counterexample and explain why it breaks the claim. Flag any step where the logic is only plausible rather than certain.

Best for: Math and formal-logic claims where a hand-wave in the middle would go unnoticed.

13. Word problem to equations

Turn on DeepThink. Translate this word problem into equations, then solve: [PASTE PROBLEM]. Define each variable in plain words, write the equations, and solve step by step. Sanity-check the answer against the real-world meaning of the problem (units, sign, plausible range) before giving the final result.

Why it works: Defining variables and unit-checking is where most word-problem errors hide; naming them forces the model to catch itself.

14. Check a proof or calculation

Turn on DeepThink. Check this work for errors: [PASTE PROOF OR CALCULATION]. Go line by line. For each step, say whether it's correct and why. If you find a mistake, point to the exact line, explain the error, and show the corrected version. If it's fully correct, say so and note the key step everything hinges on.

Best for: Grading your own homework, a spreadsheet calculation, or a colleague's derivation.

15. Probability and statistics helper

Turn on DeepThink. Help me with this probability/statistics question: [PASTE QUESTION]. State which distribution or method applies and why, show the setup and the computation step by step, and give the numeric answer with units. Then explain the result in one plain-English sentence and name one common way people get this type of problem wrong.

Why it works: Naming the method and the common trap makes the answer teachable, not just a number you have to trust.

Writing & content

Give DeepSeek a Role, a clear Task, real Context, and an exact Output format, and it matches your tone closely. For drafting, leave DeepThink off and let it write in Instant mode; turn it on only when the structure itself needs reasoning.

16. 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 read like generic AI copy.

17. 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.

18. 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.

19. 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.

Why it works: Mixing labeled angles gives you a batch of testable headlines with the reasoning attached.

20. Story or scene draft

Write a [500]-word [genre] scene. Setup: [WHO, WHERE, WHAT'S AT STAKE]. Point of view: [FIRST / THIRD], tense: [PAST / PRESENT]. Show, don't tell; let the tension build through action and dialogue rather than narration; end on a beat that pulls the reader forward. Match this voice: [DESCRIBE OR PASTE A SAMPLE]. Avoid clichés and purple prose.

Best for: Fiction drafting when you want a scene that moves instead of a summary of one.

Advertisement

Research & analysis

Turn on Search for anything current, and demand links so every claim is checkable. For documents, use the paperclip to upload a PDF, DOCX, TXT, or code file and DeepSeek reads the whole thing before you ask.

21. Summarize an uploaded document

I've attached a document. Read it in full, then give me: a 3-sentence summary, the 5 most important points as bullets, any conclusions or recommendations it makes, and anything it leaves unanswered or assumes without evidence. Ground every point in the document only; if something isn't in it, say so rather than filling the gap.

Why it works: "Ground every point in the document only" keeps DeepSeek from padding the summary with outside guesses.

22. Cited web research brief

Turn on Search. Research [TOPIC / QUESTION] and write me a brief as of this month. Cover the current state, the main viewpoints, and where things are heading. Cite every source with a link, prefer primary and recent sources, and note the publication date next to each fact. End with what's still uncertain or contested.

Best for: A current, checkable brief instead of an answer from the training cutoff.

23. Compare sources and find the gap

Turn on Search. Find the main sources on [QUESTION] and compare what they say. Build a table: Source | Key claim | Evidence given | Date | Link. Then tell me where the sources agree, where they conflict, and which claim is best supported. Flag any source that looks weak or out of date. Cite every source with a link.

Why it works: A comparison table across live sources surfaces disagreement that a single blended summary would hide.

24. Extract structured data from a file

I've attached a file. Extract [WHAT YOU NEED — e.g. every date, name, and amount] and return it as a clean table with columns [COLUMN 1, COLUMN 2, ...]. Read the whole document, don't skip sections, and leave a cell blank rather than guessing if the value isn't stated. After the table, list anything ambiguous that you had to interpret.

Best for: Pulling structured data out of a PDF, contract, or report without hand-copying.

25. Fact-check a claim with live Search

Turn on Search. 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.

Why it works: Search plus a required rating and links turns a vague "is this true?" into an auditable verdict.

Business & work

Give DeepSeek your real context and it handles the structure. Turn on DeepThink for strategy calls with real trade-offs; leave it off for straightforward drafting and formatting.

26. 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.

27. 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.

28. 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.

Why it works: The explanation plus gotchas means you can maintain the sheet yourself instead of pasting a formula you don't understand.

29. Strategy memo with trade-offs

Turn on DeepThink. Write a one-page strategy memo on [DECISION / QUESTION]. Context: [BACKGROUND, GOALS, CONSTRAINTS, DATA]. Lay out 2-3 real options, the trade-offs of each, and a clear recommendation with the reasoning. Name the key assumption the recommendation rests on and what would have to be true for it to hold. Keep it tight and executive-ready.

Best for: A memo that names the trade-offs honestly instead of arguing for a foregone conclusion.

30. 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.

Why it works: The chief-of-staff framing pushes DeepSeek to cut, not just organize, an overwhelming list.

Study & learning

DeepSeek is a strong tutor because it can reason and adapt. Turn on DeepThink when it needs to work a problem alongside you, and upload your course PDF or notes so it teaches from your actual material.

31. Explain a hard concept simply

Explain [CONCEPT] to me. First give the one-sentence core idea, then build it up with a concrete everyday analogy, then a worked example, then the common misconception people have. Assume I know [YOUR CURRENT LEVEL]. End with two questions I could answer to check I actually understood it.

Best for: Getting past a concept that a textbook explained badly.

32. Build a study plan

I need to learn [SUBJECT / EXAM] by [DATE]. I can study [HOURS] per week and my current level is [LEVEL]. Build a week-by-week plan: what to cover each week, in what order, and which topics need the most time based on how hard or foundational they are. Include checkpoints to test myself and a lighter final week for review.

Why it works: Ordering topics by difficulty and dependency produces a plan that builds, instead of a flat list of chapters.

33. Quiz me with a feedback loop

Quiz me on [TOPIC] at [LEVEL]. Ask one question at a time and wait for my answer before the next. After each answer, tell me if I'm right, correct any mistake with a short explanation, and adjust the difficulty based on how I'm doing. Keep going for [10] questions, then give me a summary of what I've got down and what to review.

Best for: Active recall practice that adapts to where you're weak.

34. Turn notes into flashcards

Turn these notes into spaced-repetition flashcards: [PASTE NOTES OR ATTACH FILE]. Make each card one clear question and one concise answer, testing a single fact or idea. Avoid cards that are too easy or that bundle several facts. Output as a two-column table (Front | Back) so I can import them, and group by subtopic.

Why it works: The one-fact-per-card rule produces cards that actually work in a spaced-repetition app.

35. Socratic tutor session

Turn on DeepThink. Be a Socratic tutor for [TOPIC / PROBLEM]. Don't give me the answer. Ask me guiding questions one at a time that lead me to work it out myself, and when I'm stuck, give the smallest hint that unblocks me rather than the solution. If I go wrong, ask a question that helps me see the mistake. Start now.

Best for: Learning to solve a type of problem yourself, not just getting this one answer.

Everyday & personal

DeepSeek is free and handles everyday planning and drafting well. Give it your real constraints, and turn on Search when the task needs current prices, hours, or availability.

36. Plan a trip with real constraints

Turn on Search. Plan a [NUMBER]-day trip to [DESTINATION] for [WHO] with a budget of [BUDGET]. We like [INTERESTS] and want to avoid [DISLIKES]. Give a day-by-day itinerary with realistic timing, group nearby stops to cut travel, and note current prices and opening hours where you can with links. Add one rainy-day backup and one splurge option.

Why it works: Grouping stops by location and adding a backup makes the itinerary usable in the real world, not just a wish list.

37. Meal plan from what you have

I have these ingredients: [LIST]. Plan [NUMBER] meals for [PEOPLE] using mostly what I have, adding as few new items as possible. For each meal give a short ingredient list, quick steps, and rough time. Respect these needs: [DIET / ALLERGIES / DISLIKES]. At the end, give me a short shopping list of only the extras I'd need.

Best for: Using up what's in the fridge without a separate grocery run.

38. Draft a difficult message

Help me write a message to [WHO] about [SITUATION]. My goal is [GOAL] and I want to sound [TONE — e.g. firm but kind]. Keep it honest and clear, don't over-apologize or ramble, and leave the door open for a good response. Give me two versions: a shorter and a slightly warmer one. Context they'll need: [BACKGROUND].

Why it works: Two versions let you pick the tone that fits, and naming the goal keeps the message from wandering.

39. Budget and money check-up

Turn on DeepThink. Here's my monthly income and spending: [PASTE NUMBERS OR CATEGORIES]. My goal is [GOAL — e.g. save X, pay down debt]. Analyze where the money goes, flag the categories that look high for my goal, and suggest a realistic adjusted budget with specific numbers. Show your reasoning and don't recommend cuts that would be too painful to stick to. This is general guidance, not financial advice.

Best for: A grounded look at your spending with realistic, stickable changes.

40. Decision journal prompt

I'm about to decide [DECISION]. Interview me to build a decision journal entry: ask me what I'm choosing, what I expect to happen, my confidence, and how I'll know later if it was right. Ask one question at a time. At the end, summarize my reasoning and expected outcome in a short entry I can revisit in [TIMEFRAME] to check how I actually did.

Why it works: Capturing your reasoning and expected outcome now is what makes hindsight useful instead of a story you rewrite later.

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

Frequently Asked Questions

Which DeepSeek model are these prompts built for?

They target DeepSeek-V3.2, the flagship released in December 2025, with a 128K-token context window and top-tier reasoning, math, and coding. They also work on earlier DeepSeek models, since DeepThink, web Search, and file upload are shared features.

What is DeepThink and when should I use it?

DeepThink is the R1-style reasoning toggle. Turn it on and DeepSeek shows its full chain-of-thought before the final answer. Use it for hard math, logic puzzles, debugging, and anything that needs maximum accuracy or an auditable reasoning trail. For quick lookups, leave it off.

How do I make DeepSeek search the web?

Turn on the Search toggle before you send the prompt. It triggers real-time web search for current events, prices, news, and anything past the training cutoff. Ask DeepSeek to cite each source with a link so the answer is checkable.

What is the text prompt formula for DeepSeek?

Role + Task + Context + Constraints + Output format. Tell DeepSeek who it is, what to do, the background it needs, the limits, and the exact shape of the answer. The golden rule: give it a hard problem plus rich context, not rigid formatting rules, and let it reason.

Is DeepSeek free and open-source?

Yes. DeepSeek is a free AI assistant at chat.deepseek.com and in the iOS and Android apps, and its models are open-source under the MIT license. You can use the chat interface at no cost and run or build on the models yourself.

Can DeepSeek read my files?

Yes. Use the paperclip to attach a PDF, DOCX, TXT, or code file and DeepSeek reads the full document. Then ask it to summarize, extract, compare, or answer questions grounded only in the attached file.

What is the difference between DeepThink and Instant or Expert?

Instant mode answers fast; Expert mode reasons more deeply. DeepThink is the reasoning layer that exposes the full chain-of-thought before the answer. Use Instant for simple tasks, and turn on DeepThink or Expert for hard problems where accuracy matters more than speed.

How do the [BRACKET] placeholders work?

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

Advertisement