# AI prompts for job search and job applications (copy-paste, 2026) Canonical: https://www.apply4me.io/en/blog/ai-job-search-prompts The best AI prompts for job search and job applications give the model three things: your facts, your target and the exact output you want. Copy the prompts below to pick target roles, pull ATS keywords from job ads, score fit, and write the application itself: a tailored cover letter, answers to form questions and a salary expectations answer. Quick answer The best AI prompt for a job search or a job application gives the model three things: your facts (CV bullets, results, tools), your target (title, level, location, salary floor) and the exact output you want (a list, a table, a 200-word letter). The best AI prompt for a job search or a job application gives the model three things: your facts (CV bullets, results, tools), your target (title, level, location, salary floor) and the exact output you want (a list, a table, a 200-word letter). Generic prompts get generic results: bland titles, irrelevant postings and cover letters that could belong to anyone. The copy-paste prompts below cover the whole process, from choosing target roles and pulling ATS keywords to writing the application itself. Both OpenAI and Anthropic say the same thing in their own prompting guides: models do better with precise instructions and the context they need. OpenAI's [prompt engineering guide](https://developers.openai.com/api/docs/guides/prompt-engineering) recommends "precise instructions that explicitly provide the logic and data required to complete the task", and Anthropic's [prompting best practices](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices) suggests treating the model as "a brilliant but new employee who lacks context". Every prompt here is built that way. --- ## How do you write an AI prompt for job search? Use this 3-part input formula in every prompt: 1. **Your profile (facts):** role history, hard skills, tools, results, industries. 2. **Your target (constraints):** titles, level, company type, work style, salary floor, location or time zone. 3. **Your fit rules (non-negotiables):** tech stack, team size, travel limit, visa needs, mission. ### Quick prompt template (copy/paste) > Act as a job search strategist. Based on my background below, propose (1) 8 realistic target job titles, (2) the top 20 ATS keywords to include, (3) a Boolean search string for LinkedIn and Google Jobs, and (4) red flags to avoid. > **Background:** [paste 6-10 bullets] > **Constraints:** [level, location, salary, remote/hybrid, industries] > **Non-negotiables:** [3-5 items] > **Output format:** bullets + one table of keywords grouped by category. --- ## Prompt for a job application: cover letter and form answers These three prompts write the parts of an application that a recruiter actually reads. Paste the full job ad every time; a prompt without the ad can only produce a generic letter. ### Prompt A: cover letter for one specific job > Write a cover letter for the job ad below. Rules: > > - 180-230 words, 3 short paragraphs, no "I am writing to apply" opening > - Paragraph 1: why this company and this role, using one specific detail from the ad > - Paragraph 2: two results from my background that match the two most important requirements, with numbers > - Paragraph 3: one sentence on availability and a clear next step > - Use the ad's wording for skills, but do not copy whole sentences > - Do not invent anything that is not in my background > **Job ad:** [paste] > **My background:** [paste CV or 8-10 bullets] > **Language and tone:** [English, plain and direct] For more cover letter variants, see [an AI cover letter generator that sounds human](https://www.apply4me.io/en/blog/ai-cover-letter-generator-that-sounds-human). ### Prompt B: answers to application form questions > I am filling in the online application form for the job below. For each question, write an answer that is (1) true to my background, (2) under the character limit I give, and (3) uses the ad's keywords where they fit naturally. For yes/no knock-out questions, tell me which answer is accurate for me and flag any answer that could screen me out. > **Job ad:** [paste] > **My background:** [paste] > **Form questions and limits:** [paste each question with its character limit] Knock-out questions decide many applications before a person reads them; see [how ATS knock-out questions work](https://www.apply4me.io/en/blog/ats-knock-out-questions) and [how to answer LinkedIn Easy Apply questions](https://www.apply4me.io/en/blog/linkedin-easy-apply-questions-how-to-answer). ### Prompt C: "Why do you want to work here?" and salary expectations > Write two short answers for this application. > > 1. "Why do you want to work here?" in 60-80 words: one specific fact about the company from the text below, one link to my experience, no flattery. > 2. "What are your salary expectations?" as a range, based on the range in the ad if there is one, or on the market data I paste. Add one sentence that keeps the range open to the full package. > **Job ad and company page text:** [paste] > **My background:** [paste] > **Market data or my target range:** [paste] Before you type a number, read [how to answer salary expectations on applications](https://www.apply4me.io/en/blog/how-to-answer-salary-expectations-on-applications). --- ## AI prompts to clarify your target roles If you apply to "anything in marketing", the AI mirrors that vagueness. These prompts narrow your target so your searches and CV tell one story. ### 1) Role clustering prompt (find your best-fit lanes) > Review my resume summary below and group my experience into 3-4 career lanes (e.g., Growth Marketing, Lifecycle, Product Marketing). For each lane, list: ideal job titles, typical KPIs, common tools, and the industries where my background is strongest. Then recommend the single best lane for the next 90 days and explain why. > **Resume summary:** [paste] **How to use it:** pick one lane for 60-90 days. Searches and keywords work better when they stay consistent. ### 2) Transferable skills to target titles (for career changes) > I want to move from [current role] to [target function]. Map my transferable skills to that function using the language recruiters use. Give me: > > - 10 target titles (entry/mid/senior options) > - 10 bridge roles that are easier to land first > - 15 keywords and 8 tools to learn > - 5 portfolio or project ideas that prove capability in 30 days ### 3) Two-sentence positioning prompt (LinkedIn headline and summary) > Write 5 versions of a 2-sentence positioning statement for my LinkedIn About section. Each must include: my target role, specialty, measurable results, and the types of teams or industries I help. Keep each under 55 words and avoid buzzwords. > **Background:** [paste] --- ## Which keywords should you use? Prompts to extract ATS keywords from job ads The goal is not to cram keywords. It is to use the right ones and back each one with proof: a number, a scope, a tool. ### 4) Keyword extraction and priority prompt (from 5 job ads) > Analyze these 5 job descriptions for the same target role. Create: > > 1. a ranked list of the top 25 keywords and phrases by frequency and importance > 2. a keyword-to-proof map: how I should show each keyword with an example bullet > 3. a short list of keywords I should NOT use because they overstate my experience > **Job descriptions:** [paste 3-5] > **My background:** [paste 6 bullets] ### 5) Skills gap prompt (turn missing keywords into a plan) > Compare my current skills to the job requirements below. Identify: > > - must-have gaps vs nice-to-have gaps > - the fastest way to close each gap (course, project, certification, work equivalent) > - a 2-week mini-project I can do to credibly claim the skill > **Job description:** [paste] > **My resume bullets:** [paste] ### 6) Resume bullet rewriting prompt (ATS-friendly, still human) > Rewrite these resume bullets for the job description below. Keep them truthful, specific and results-based. Use strong verbs, include numbers where possible, and align to the top keywords without copying the job description word for word. Give 2 versions: concise (1 line) and detailed (2 lines). > **Job description:** [paste] > **My current bullets:** [paste] For a full tailoring workflow, see [AI resume tailoring prompts for each job description](https://www.apply4me.io/en/blog/ai-resume-tailoring-prompts-for-each-job-description) and the [AI resume summary generator prompt](https://www.apply4me.io/en/blog/ai-resume-summary-generator-prompt). --- ## How do you judge job fit before applying? Prompts to screen ads like a recruiter Better matches come from saying no faster. ### 7) Fit scoring prompt (with a rubric) > Create a fit-scoring rubric (0-100) for this job based on: skills match, tool match, domain match, seniority, results/KPIs and work style. Then score me using my background. Give: > > - overall score + sub-scores > - top 5 reasons I'm a strong match > - top 5 risks > - whether to apply now, apply after changes, or skip > **Job description:** [paste] > **My background:** [paste] ### 8) Hidden requirements prompt (decode vague ads) > Read this job description and infer what the hiring manager most likely wants (team context, urgency, pain points, partners, KPIs). List 10 questions I should ask in a recruiter screen. > **Job description:** [paste] ### 9) Level and pay reality check prompt > Based on the responsibilities and required skills in this job description, estimate the most likely level (mid/senior/lead) and a typical pay range in [location]. Explain which signals you used (scope, ownership, team leadership, system complexity). > **Job description:** [paste] > **Location:** [city/country] *Note:* an AI estimate is a starting point. Check it against real ads: in Poland only 49.4% of 20,543 active ads stated a salary on 24 October 2026 ([Apply4Me job market report](https://www.apply4me.io/en/job-market-report)), so compare with the ads that do. --- ## A 30-minute workflow with these prompts ### Step 1 (5 minutes): pick one target lane Use Prompt 1. Write down 3 target titles, 2 adjacent titles and, optionally, 1 industry. ### Step 2 (8 minutes): build your keyword bank Paste 3-5 job ads into Prompt 4 and keep the top 25 keywords, the tool list and the KPI language. ### Step 3 (7 minutes): create Boolean searches > Create 4 Boolean search strings for LinkedIn and Google Jobs using my target titles and keyword bank. Include one string for remote roles and one for my local area. Exclude irrelevant roles (e.g., internships, senior director). > **Target titles:** [list] > **Keywords:** [paste] > **Location:** [city/remote] > **Exclude:** [terms] On LinkedIn, type AND, OR and NOT in capitals; see [LinkedIn job search filters](https://www.apply4me.io/en/blog/linkedin-job-search-filters-best-settings). ### Step 4 (5 minutes): pre-screen ads before you apply Score the top 5 roles with Prompt 7: - **80+**: apply now - **65-79**: apply if you can tailor quickly - **under 65**: skip or save for later ### Step 5 (5 minutes): write the application Use Prompt 6 for your bullets, then Prompts A, B and C for the cover letter and form answers, and apply. --- ## ChatGPT, Gemini, Claude or Apply4Me: which tool for which step? | Tool | Best for | Strengths | Limits | Use it for | | --- | --- | --- | --- | --- | | ChatGPT | Prompt-driven strategy and rewriting | Strong iterative editing, flexible output | You supply the structure; drifts into generic wording | Keyword banks, bullet rewrites, fit scoring | | Gemini | Research workflows (depends on your setup) | Summarizing companies and role themes | Results depend on sources | Company lists, role research | | Claude | Long documents | Reads many job ads and a full CV at once; good tone control | Still needs a system for tracking | Multi-ad keyword analysis, cover letters | | Apply4Me | Running the search | Offers from 50+ job boards on one list with a match score, a tailored CV and cover letter per offer, application tracking, sends only after you approve | Not a general writing assistant | Turning your prompts into a steady flow of applications | *Disclosure: Apply4Me publishes this article and is one of the tools in the table.* **Verdict:** use a general AI assistant for thinking and writing (titles, keywords, fit rubrics, letters), then a dedicated tool to apply and track at a steady pace. --- ## Advanced prompts: networking, recruiter screens and interviews ### 10) Networking message prompt (short and specific) > Write 3 versions of a LinkedIn message to a [role] at [company] about [specific team/product]. Each message must: > > - be 60-90 words > - mention one specific detail from the company page or their profile > - ask one clear question > - avoid flattery and "I'd love to connect" > **My background:** [paste] > **Target role:** [paste] > **Company detail:** [paste] ### 11) Recruiter screen prep prompt > I have a 20-minute recruiter screen for [role]. Create: > > - a 30-second "tell me about yourself" > - 5 likely questions + my best answers based on my background > - 5 questions I should ask that show seniority > - a salary expectation answer that keeps my options open > **Job description:** [paste] > **My background:** [paste] ### 12) Interview story bank prompt (STAR) > Build a story bank of 8 interview stories mapped to: leadership, ambiguity, stakeholder management, conflict, execution, analytics, customer focus and innovation. For each story: a 3-bullet STAR outline + numbers + what I learned. > **My experience bullets:** [paste] --- ## Common mistakes that make AI job search prompts fail ### Mistake 1: asking for "the best jobs for me" with no constraints **Fix:** give level, location, salary floor, industries and 5-10 concrete experience bullets. ### Mistake 2: treating one job ad as the whole market **Fix:** use 3-5 ads for the same role, then extract patterns with Prompt 4. ### Mistake 3: copying the job ad into your CV or letter **Fix:** mirror the keywords, but prove them with results: scale, scope, tools and numbers. ### Mistake 4: not tracking what works **Fix:** track roles, response rate and interviews per CV version, then change one thing at a time. --- ## Sources - OpenAI, [Prompt engineering guide](https://developers.openai.com/api/docs/guides/prompt-engineering), checked 25 October 2026. - Anthropic, [Prompting best practices](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices), checked 25 October 2026. - Apply4Me, [Polish job market report, Q4 2026](https://www.apply4me.io/en/job-market-report) (data as of 24 October 2026). ## Frequently Asked Questions What is the best AI prompt for a job application? One that includes the full job ad, your real background and strict output rules (length, structure, no invented facts). Prompt A above writes a 180-230 word cover letter; Prompt B answers form questions within their character limits. What are the best AI prompts for job searching? Start with three: a role clustering prompt to pick target titles, a keyword extraction prompt run on 3-5 job ads, and a fit-scoring prompt to decide which ads deserve an application. How many job descriptions should I paste into AI for keyword analysis? Three to five ads for the same role and level. That shows the skills that repeat without overfitting to one company's wording. Will AI-written applications get flagged by ATS or recruiters? An ATS parses structure and keywords; it does not grade writing style. Recruiters do notice generic phrasing, so keep the language plain and back every keyword with a fact from your experience.