ChatGPT Review
The universal job-search tool that's free, fast, and only as good as the prompt you feed it.
✓ Pros
- • Free tier handles 90% of job-search use cases without paying a penny
- • Massive prompt flexibility means you can attack any CV, cover letter, or interview problem
- • Instant rewrites and iterations beat waiting for a template tool to regenerate
- • Works for any job title or industry, from graduate retail to senior engineering
- • Custom GPTs let you build repeatable workflows for specific tasks like CV bullet rewriting
✗ Cons
- • Hallucinates metrics and achievements if you don't anchor it to your real experience
- • Default output voice is recognisably AI and will get flagged by any recruiter paying attention
- • Won't push back on weak material unless you specifically ask it to critique you
- • No formatting or template output, you get plain text and have to lay it out yourself
- • Requires actual prompting skill to get results better than a dedicated tool would produce
Every single AI resume tool I’ve reviewed is a ChatGPT wrapper with a prettier UI. That’s not cynicism, that’s literally true. The £29-a-month CV builder you’re thinking about subscribing to is, under the bonnet, sending your input to OpenAI’s API and returning the response inside a nicer interface.
So the question I keep getting asked by candidates is: do I need the wrapper, or can I just use ChatGPT directly?
Here’s the honest answer from 12 years of reading CVs. If you can prompt well, you don’t need the paid tools. If you can’t, ChatGPT’s raw output is worse than a cheap template because at least the template stops you doing something stupid. This review is about working out which camp you’re in.
What ChatGPT actually is (for job seekers)
Stripped of the hype, ChatGPT is a conversational text generator you can point at almost any writing problem. For a job seeker that means it can rewrite, restructure, summarise, critique, and role-play, all inside a chat window.
It has no idea who you are. It doesn’t know your achievements, your company’s revenue, or whether you actually led a team of four or a team of forty. Everything useful it produces comes from what you put in. That’s the thing candidates misunderstand most. They ask it to “write my CV” and then copy whatever comes back. What comes back is a fictional average of every CV on the internet, lightly seasoned with the two sentences you typed.
Use it as a co-writer, not a ghostwriter, and it becomes genuinely useful.
The five ChatGPT job-search use cases that actually work
1. CV bullet rewriting
This is where it earns its place. Paste one weak bullet, ask it to rewrite the bullet five different ways with varied structure, and pick the one that matches your voice. You keep the facts, it improves the sentence.
Here's a bullet from my CV: "Responsible for managing a team and hitting sales targets."
Rewrite it in 5 different versions. Keep every factual claim identical, but vary the opening verb and sentence structure. No buzzwords. British English. Each version under 20 words.
2. Cover letter drafting
Paste the job advert, paste your CV, and ask for a first draft. Never ship the first draft, but it gets you past the blank page in 30 seconds. I’d say 60% of cover letter friction is starting, and ChatGPT removes it.
3. Interview Q&A practice
Feed it the job description and ask it to generate the 10 most likely interview questions for that role. Then go one by one and ask it to critique your draft answer. It’s a patient practice partner that never gets bored at 11pm.
You are a hiring manager interviewing for [role title] at [industry]. Based on this job spec [paste spec], generate the 10 interview questions I'm most likely to face. Include 3 competency questions, 3 technical, 2 about culture fit, and 2 curveballs.
4. LinkedIn profile polish
The About section is where most profiles fall apart. Paste yours, ask ChatGPT to critique it as if it were a recruiter skimming for 8 seconds, then ask for three rewritten versions at different tones (direct, warm, sharp). Pick the closest to your voice and edit from there.
5. Career research
Changing industries or seniority level? ChatGPT is fast at summarising what a day looks like in an unfamiliar role, what the typical career path is, and what the interview process usually involves. Treat the output as a starting point for proper research, not the final word, because it gets details wrong.
The three use cases where ChatGPT disappoints
Full CV generation from scratch. Ask it to write you a full CV and you get four paragraphs of management consultancy voice with invented metrics. This is where hallucination is worst because the model is filling in gaps with plausible-sounding nonsense. I’ve seen candidates send CVs claiming they “drove a 34% increase in stakeholder engagement” when no such metric exists in their actual job.
Salary negotiation scripts. It’ll give you textbook advice that any recruiter hears coming a mile off. “I appreciate the offer and based on my research and market rates…” will make the hiring manager’s eyes glaze over. Negotiation needs your specific context and relationship with the company, not a generic script.
Formatting. ChatGPT outputs plain text. If you want a CV you can actually send, you’re pasting the content into Word, Google Docs, or a template tool. It won’t save you the layout work.
The prompting skill gap is the whole game
This is the bit nobody wants to hear. The difference between a candidate who gets a shortlist using ChatGPT and one who gets binned is about 20 minutes of prompting practice.
I can tell within 10 seconds when someone has pasted ChatGPT output straight into a cover letter. The giveaways: “I am writing to express my keen interest”, “my unique blend of skills”, three-part lists where every item starts with an -ing verb, and metrics that are suspiciously round (25%, 30%, 40%).
The candidates who use ChatGPT well are the ones who treat it like a junior copywriter. They give it constraints (“no buzzwords, 20 words max, British English, first person”), they reject the first draft, they iterate, and they edit the final output in their own words. The candidates who use it badly ask “write me a CV” and hit send.
The paid CV tools work around this gap by hard-coding constraints into their prompts. That’s what you’re actually paying for. If you can write those constraints yourself, you don’t need them.
Free vs Plus: do you need $20/month?
For a single job search lasting 4 to 12 weeks, the free tier is genuinely enough for most people. You get GPT-4o with limited messages per 3-hour window, which resets quickly. For CV bullets, cover letters, and interview prep, you’ll rarely bump into the cap.
Plus at $20/month is worth it if any of these apply: you’re hitting the free-tier cap during focused prep sessions, you want voice mode for spoken interview practice, you want priority access during peak times, or you want to build your own Custom GPTs for repeated workflows.
For 80% of job seekers I’d say stay free. Run one focused week of free-tier use, and only upgrade if you’re actually hitting the limits.
ChatGPT vs dedicated job-search tools
vs Teal. Teal wins on keeping track of applications, job tracker, and CV version management. ChatGPT wins on raw writing flexibility. If you’re applying to 30+ roles and losing track of who you’ve sent what, Teal is worth it. If you’re sending 5 targeted applications, ChatGPT plus a spreadsheet is fine.
vs Rezi. Rezi is basically ChatGPT with hard-coded ATS constraints and a template output. If you trust yourself to write ATS-friendly bullets in ChatGPT, you save $29/month. If the phrase “ATS-friendly” makes you nervous, Rezi removes that decision for you.
vs Grammarly. Different tools. Grammarly catches grammar and tone issues in polished drafts. ChatGPT generates and rewrites. Use both, not either/or. Grammarly Premium is genuinely useful as a final pass before sending.
The pattern: dedicated tools beat ChatGPT when the problem is workflow and constraints. ChatGPT beats them when the problem is raw writing ability.
My verdict
ChatGPT is the best job-search AI tool on the market if you can prompt it well, and a liability if you can’t.
FAQs
Can recruiters detect ChatGPT-written CVs and cover letters? Yes, and faster than candidates think. The default ChatGPT voice has signature tells: over-formal openers, three-part lists, suspiciously round metrics, and phrases like “proven track record” and “results-driven”. If you paste output without editing, experienced recruiters catch it in 10 seconds. Edit it into your own voice and nobody can tell.
Is the free tier enough for a whole job search? For most people, yes. Unless you’re running full interview role-plays for hours at a time, you’ll stay under the free-tier limits. The one exception is if you want voice mode for spoken interview practice, which is Plus-only.
Does Plus at $20/month actually help you get more interviews? Not directly. The Plus tier doesn’t produce better CVs than free GPT-4o. It helps if you want features like voice mode, Custom GPTs, and priority access. Don’t pay $20 thinking it’ll get you a job faster, pay it if you’ll use the specific features.
Are Custom GPTs worth using for job search? Yes, for repeated tasks. Building a Custom GPT with permanent instructions like “always produce 3 variants, British English, no buzzwords, under 20 words” saves retyping those constraints every time. The community-built ones (CV Reviewer, Interview Coach) are a mixed bag, some good, most mediocre. Build your own or stick to direct prompting.
What about Claude or Gemini, are they better? Claude often produces more natural writing and is my personal go-to for cover letters. Gemini integrates with Google Docs, which some people find useful. ChatGPT has the biggest ecosystem of Custom GPTs and the strongest voice mode. For most job seekers the differences don’t matter, pick whichever you’re comfortable with and learn to prompt it properly.
Related reading
- Can recruiters tell if you used AI? — the 8 dead giveaways and how to use ChatGPT without getting caught
- ChatGPT prompts for resume writing — the prompts I’d actually use
- AI resume buzzwords recruiters hate — what to strip from any AI output
- Teal review — when a dedicated CV tool beats ChatGPT
- ChatGPT vs Teal comparison — the head-to-head on tracker features, per-role tailoring, and raw prompt flexibility
- Best AI resume builders 2026 — how the paid tools compare
How to actually use ChatGPT for your job search
Three steps, in this order.
Step one: set your constraints once. Open ChatGPT, go to Settings, Personalisation, Custom Instructions. In the “What should ChatGPT know” box, paste a short summary of your experience and target roles. In the “How should ChatGPT respond” box, paste your rules: British English, no buzzwords, never invent metrics, always produce 3 variants, ask me before making assumptions.
Step two: never accept the first draft. Whatever it produces, reply with “make it more specific”, “make it shorter”, or “rewrite in my voice based on this sample: [paste your actual writing]”. The second draft is almost always better than the first.
Step three: edit in your own words before you send. No exceptions. The final version should have sentences you wrote, not sentences the model wrote. That’s the difference between a CV that gets shortlisted and one that gets binned.
Best for
- → Anyone who can write a clear prompt and iterate on feedback
- → Budget-conscious job seekers who refuse to pay $29/month for a CV builder
- → Candidates in unusual industries where template tools have no relevant data
- → Late-career professionals comfortable rewriting their own material with AI assistance