# AI Slop Detector: Does Your Website Look AI-Generated? | Koso > Free AI slop detector for landing pages. It renders your page in a real browser and checks 34 tells of AI-built design and copy. Every finding is boxed on a screenshot, and you get a score out of 100 plus a fix prompt. [](/) [Services](/#services)[AI OS](/ai-os)[Training](/training)[Education](/education)[Agents](/agents)[Learn](/learn) … [Book a call ](/book) 1. [Home](/) 2. [AI Agents](/agents) 3. AI Slop Detector Free AI agent · built by Koso # Is Your Site AI Slop? Paste a page and we'll open it in a real browser and measure what it actually drew. You get a score out of 100 for how much of it is template default, with every finding boxed on a screenshot. Scan page Try:linear.appkoso.aistripe.com Free · first run without sign-in · results in \~30 seconds ## Want An Agent Like This For Your Business? This is a real browser, a set of code checks and a decision model working as one pipeline. Koso builds these for clients, pointed at their own documents and inboxes. [Book a free call ](/book)[All agents ](/agents) ## We Measure The Page Itself. A screenshot can't tell you the font is Inter or that your grey text fails contrast. The browser can, so that's where most of the checks run. Koso AI only gets the questions that need judgment, like whether a section could sit on anyone else's site. true: scanned in the last day, answered instantlyfalseresult saved for 24 hourspage URLyou paste it aboveIFcache checkscanned in last 24h?page-capturereal browser, or HTMLmeasurecomputed styleskoso-ai-judgeeach section, the copyscorefixed formulaslop reportindex · boxes · fix promptCloudflare browser{ }pure code, no AIAIKoso AI{ }pure code, no AIresults cache Swipe the diagram sideways to follow the full flow. The actual pipeline behind every scan. Dashed wires are the tools each node uses. Only one node involves a model, and it answers with probabilities, not prose. ## Everything We Check. 34 tells, each with a weight. The number next to each one is how many points it adds when it's fully there. 4 of them are judged by Koso AI; the rest are measured in code. ### Colour and effects * +3Purple gradient. A gradient running through purple, violet or indigo. It's the house colour of AI-built pages. * +2Gradient headline text. Headline words filled with a colour gradient instead of one colour. * +2Purple buttons. Filled buttons in indigo or violet, the colour most AI builders reach for first. * +2Glows and blurred blobs. Blurred coloured orbs and coloured shadows used as decoration. * +2Frosted glass panels. Translucent cards with a blurred background showing through. * +2Dark mode with neon accents. A near-black page lit up with purple or cyan accents. * +1Graph-paper background. A faint grid or dot pattern sitting behind the content. * +1Default cream background. A warm off-white page. Anti-slop advice made it the new default in 2026. * +2Washed-out body text. Body text that fails the WCAG AA contrast ratio of 4.5 to 1. ### Type * +2Default font. Inter or a system font for both the headlines and the body. * +1This year's fashionable font. A typeface that's everywhere right now, like Space Grotesk or Instrument Serif. * +1Italic serif accent word. One or two italic serif words dropped into a sans-serif headline. * +1Squeezed headline. Headline letters pulled tighter than -0.04em, the tracking-tighter default. * +1Tiny all-caps labels. Small spaced-out capital labels above section after section. ### Layout * +2Badge above the headline. A small rounded pill above the main headline, usually announcing something new. * +2Centred hero stack. Headline, subheading and buttons all centred in one column. The stock hero. * +2Icon-on-top feature cards. Feature cards that each open with an icon in a small rounded square. * +1Three identical cards in a row. Exactly three cards of the same size and structure side by side. * +2Bento grid. Features packed into rounded tiles of mismatched sizes. * +1Coloured stripe on cards. Cards or callouts with a thick coloured border down one edge only. * +1Row of big round numbers. A banner of large figures like 10k+ and 99.9%. * +1One, two, three steps. A numbered walkthrough, whether or not the product needs one. * +1Grey logo wall. A strip of greyed-out company logos under the hero. * +1Emoji as icons. Emoji standing in for icons, bullets or section markers. * +1Sparkles for AI. A sparkle icon or ✨ used to say that something involves AI. * +1Things that never stop moving. Marquees, pulsing dots and other animations on an endless loop. * +2Sections straight from a template. Most sections could sit unchanged on thousands of other sites.Koso AI ### Copy * +1Em dashes everywhere. Heavy use of the long dash, a habit of AI writing tools. * +2Buzzwords. Three or more stock words like seamless, unlock, supercharge or elevate. * +3Placeholder text left in. Leftovers like lorem ipsum or John Doe. * +2Headline doesn't say what it is. A headline a reader can't turn into 'it does X for Y'.Koso AI * +1Slogans in threes. Punchy three-part lines like 'Build. Ship. Scale.' used more than once.Koso AI * +1'Not just X, it's Y'. Defining the product by what it isn't, for emphasis.Koso AI ### Finish * +1No share image or favicon. Links to the page share without a preview image, or the tab has no icon. ### Signs a person made it * −4A typeface with character. The headline or body font isn't a default or this year's trend. * −4Shows the actual product. Screenshots, recordings or demos of the real interface. * −4Specific copy. Real customer names and numbers with context. * −4Its own layout. The page doesn't follow the standard hero, logos, three cards, testimonials, pricing order. ### What the index means * Clean (0 to 19): This page looks like somebody designed it. * A few tells (20 to 39): Mostly considered, with a couple of defaults showing. * Templated (40 to 59): Half template, half decisions. * Heavy slop (60 to 79): Most of this page came straight out of the template. * Pure slop (80 to 100): Almost every choice on this page is a default. Weighted points are capped, so a page can't score past 100 by piling up small tells. The final index blends them with Koso AI's read of how template-like the whole page is. ## AI Slop Detector FAQ ### What is AI slop in web design? It's the look a page gets when nobody made the design decisions: the builder did. Lovable, v0, Bolt and coding agents without a design system all reach for the same statistically safe defaults, mostly shadcn and Tailwind starters, so their pages share a face. A purple gradient, a pill badge above a centred headline, three identical cards with an icon on top, a grey logo strip, Inter everywhere. Any one of those is fine. All of them together tells a visitor that the page came out of a template, and they start reading everything else on it with that in mind. ### How does the detector check my page? It opens your page in a real Chrome browser on Cloudflare, scrolls it so lazy sections load, and reads the computed style of every element: the colours the browser actually painted, the fonts that loaded, what sits above the headline, how the cards are laid out, the contrast of your body text. That's where most of the checks are decided, by code. Then Koso AI reads a plain-English description of each section and your copy and answers the questions code can't: is this section stock, does the headline say what the product does, which tool probably built it. ### Is the score made up by an AI? No. The score is a fixed formula: each tell found adds its weight, Koso AI's view of how template-like the whole page is counts for 40%, and each sign that a person made deliberate choices takes points off. Koso AI doesn't write anything; it returns probabilities, and a tell only counts when it's at least 65% sure. Every finding on the page shows its evidence, and the screenshot boxes where it was measured, so you can check the work. ### What is Koso AI? It's the decision model behind the judgment calls. Instead of writing text it answers typed questions (yes or no, pick one, score this) with calibrated probabilities, which is what this job needs: a judgment we can threshold, not a paragraph. It reads text only, so the browser does the looking and Koso AI gets a plain description of each section. ### The scan says my site shows a bot check. What now? Some sites answer automated browsers with a CAPTCHA or an error page. We don't try to get past those, that's your site's call. Instead, open your site in your own browser, paste the one-line snippet we give you into the browser console, and paste the result back here. It runs exactly the same measurements on the page you can see, then scores it the usual way. The only thing missing is the screenshot overlay. ### My brand really is purple. Is that a problem? Not on its own. Purple is one tell out of 34, worth a few points. The detector flags it because it's the single most common colour on AI-built pages, so a visitor's first impression leans that way. What pushes a score up is stacking: purple plus the gradient headline plus the glow plus the badge plus the icon cards. Keep the purple, drop the defaults around it, and the score falls. ### How is this different from Slop Doctor and other slop checkers? Slop Doctor is great fun and the idea behind this one: it screenshots the page, has a vision model describe it and asks a decision model about the description. This detector measures the rendered page directly instead of describing a picture of it. So it knows the font is Inter rather than guessing, and it catches things a screenshot hides, like contrast ratios and letter-spacing. Every finding is boxed on the screenshot, signs of a human hand count in your favour, and you leave with a prompt you can paste into Lovable, Cursor or Claude Code. ### Does a high score mean my page is bad? No. It means your page looks like a lot of other pages. Plenty of high-scoring pages convert well, and a template is a sensible way to launch. The score is about distinctiveness: if a visitor has seen your hero on ten other sites this month, they'll give your product less of the benefit of the doubt. Treat the fix list as a way to look like yourself, not as a quality grade. ### Is it free, and how often can I run it? Free. Your first scan needs no sign-in; after that we ask for a name and email, no password, no credit card. Up to eight scans an hour from one address, and the same page scanned again within 24 hours is answered instantly from the saved result. The browser time comes from Cloudflare's free tier, so on a very busy day a scan may run on the page's HTML alone, without a screenshot. The report says when that happens. ## Try The Other Agents. Every tool here is a live multi-agent pipeline, free to run, reasoning on display. [All agents ](/agents) [LiveDomain AppraisalSix agents value any domain: real web history, a 1000-point score, real sales, retail and wholesale prices.Appraise a domain ](/agents/domain-appraisal)[LiveCompany ResearchSix agents build a dossier on any company and benchmark it against its rivals.Research a company ](/agents/company-research)[LiveAI Agent ReadinessHow well your website works when ChatGPT and Perplexity come to read it, out of 100.Check your site ](/agents/ai-readiness)[LiveProduct Market FitA manager agent hires analyst agents to test whether your product fits its market.Analyze your product ](/agents/product-market-fit)[LiveAgent Workflow GeneratorDescribe the agent you want. It asks 2 or 3 sharp questions, then draws the workflow and writes the build spec.Design your agent ](/agents/workflow-generator)[LiveICP GeneratorReads your real pricing page and job board, then builds the ideal customer profile: firmographics, the buying committee, triggers, and who to say no to.Build your ICP ](/agents/icp-generator)[LiveProspect ResearchGive a name and a company. Agents read the public professional web and write your call-prep note: who they are, how to talk to them, and what to open with.Research a prospect ](/agents/prospect-research)[LiveAI Skill BuilderDescribe a capability and it writes you an installable SKILL.md for Claude Code, Cursor and Copilot, after showing you what already exists on skills.sh.Build a skill ](/agents/skill-builder)[LiveCan I Run This LLMReads your graphics card from the browser, then measures every popular open model against it: what runs, how fast, and the command to pull it.Check your PC ](/agents/can-i-run-this-llm)[Coming soonBuild Your Own AgentSoon you'll design an agent like these yourself, right here. Sign in and you're on the list for early access.Get on the list ](/login) A custom AI development agency. Bespoke agents, copilots and products engineered for teams that want a moat, not a subscription. Ideanab Private Ltd is a company registered in England and Wales, no. [16344048](https://find-and-update.company-information.service.gov.uk/company/16344048). Registered office: 124 City Road, London, EC1V 2NX, United Kingdom. ### Company * [Services](/#services) * [Custom AI Agency](/custom-ai-agency) * [AI Product Development](/ai-product-development) * [AI OS](/ai-os) * [AI Engineering](/ai-engineering-company) * [AI Consulting](/ai-consulting-company) * [AI Consulting Services](/artificial-intelligence-consulting-services) * [Training](/training) * [Education](/education) * [Free AI Agents](/agents) * [Learn AI](/learn) * [Fractional CAIO](/caio) * [About Us](/about) * [The Koso Deck](/deck) * [Careers](/careers) ### Contact * [Contact us](/contact) * [Book a call](/book) * [LinkedIn](https://www.linkedin.com/company/koso) * [Instagram](https://www.instagram.com/okaashish) ### Legal * [Terms of Service](/terms) * [Privacy Policy](/privacy) * [Cookie Policy](/cookies) * [Acceptable Use Policy](/acceptable-use) © 2026 Koso, a brand of Ideanab Private Ltd. All rights reserved. [Terms of Service](/terms)[Privacy Policy](/privacy)[Cookie Policy](/cookies)[Acceptable Use Policy](/acceptable-use) Crafted with care · Built to compound