Drop a URL. Screenshot, dissect, deliver the review your team lead wouldn't.
How it works:
1. Screenshot: We take full-page screenshots at desktop (1440px) and mobile (390px) — including content below the fold that most auditors never look at. 2. Analyze: A vision-capable model grades Design, UX, and CRO against a rubric built from first principles — not generic SaaS critique templates. 3. Roast: Each callout names the exact element, the principle it violates, and the one change that would 10× that pillar. Plus one quotable kill shot.
What does it actually audit? Design (visual hierarchy, typography, contrast, spacing), UX (navigation, forms, cognitive load, mobile usability), and CRO (headline clarity, CTA placement, trust signals, funnel friction). Each category is scored 0–100 and graded like a report card.
Is the roast grounded in what's on the page? Yes — that's the whole point. Every callout must reference something actually visible in the screenshot. No generic 'improve your typography' advice.
Will anyone else see my roast? No. Roasts are private by default. Only you can access your result. There's no public slug or shareable URL — you share by downloading the result card.
How long does a roast take? Usually 30–90 seconds. Screenshots take the longest. AI analysis runs in parallel. Free roasts run synchronously; authenticated roasts queue in the background.
What if my roast fails to generate? If your roast fails due to a technical issue — bot-blocked site, blank screenshot, or a pipeline error — your credit is automatically returned. No request needed. If the roast generated but you disagree with the analysis, credits are not refunded: the AI reflects what was on the page at screenshot time.
1 review
Credits are non-refundable. Secure checkout by Razorpay.
This app is free to use.
