Parallax

Real-time epistemic annotation — detect rhetorical framing, selective evidence, and persuasion techniques in any content.

Read the bias. Write without it. Parallax is an AI-powered critical reading tool that detects rhetorical framing, logical fallacies, and persuasion techniques in anything you read or watch online. Click the icon on any article or YouTube video, and Parallax scans the content in seconds — highlighting exactly where and how the author is shaping your perception. This is not a political bias detector. Parallax identifies specific epistemic techniques that bypass critical thinking, regardless of which side they come from. What Parallax detects: Parallax identifies 26 rhetorical techniques across five categories: Framing: Selective Framing, Omission of Context, Oversimplification Evidence Manipulation: Cherry-Picked Evidence, Anecdotal Evidence Rhetorical Devices: Loaded Language, Appeal to Emotion, Motte and Bailey, Bandwagon Effect, Gish Gallop Logical Fallacies: False Equivalence, Appeal to Authority, Straw Man, Ad Hominem, Red Herring, Slippery Slope, Hasty Generalisation, Circular Reasoning, False Dichotomy, Moving the Goalposts, Whataboutism, Appeal to Nature Cognitive Biases: Anchoring, Survivorship Bias, Availability Heuristic, Confirmation Bias Each detection includes a confidence rating, a detailed explanation of why the passage may mislead, a counterargument with specific missing context, and academic sources you can verify yourself. How it works: 1. Navigate to any article, news story, opinion piece, or YouTube video 2. Click the Parallax icon in your toolbar (or right-click and select "Analyze with Parallax") 3. The sidebar opens with a real-time scanning animation 4. In seconds, highlighted passages appear in the article and a full analysis loads in the sidebar 5. Click any highlight to see the technique, explanation, and counterargument 6. Browse the Insights tab for a complete overview with your Framing Score (0-100) Works everywhere you read: News articles (BBC, NYT, CNN, Reuters, The Guardian, and thousands more) Opinion pieces and editorials Blog posts and newsletters YouTube videos (full transcript analysis with timestamp navigation) Twitter/X threads Any web page with text content Key features: Framing Score: Every piece of content gets a score from 0 to 100, showing how heavily rhetorical techniques are used overall. Scores are broken down by category with a visual chart. Inline Highlights: Detected passages are highlighted directly in the article with colour-coded underlines by category. Hover for a quick summary, click for the full analysis. Detailed Analysis: Each insight includes a plain-language explanation of why the technique matters, what context is missing, and what to look for next time. Academic sources are provided where relevant. Technique Glossary: A built-in reference guide with definitions, examples, and explanations for all 26 techniques. Learn to spot them yourself. Blind Spots: An AI-powered analysis of what the article doesn't say — perspectives, evidence, and context that were omitted entirely. YouTube Integration: Full transcript extraction with timestamp-synced annotations. Click a timestamp to jump to that moment in the video. Annotations appear as overlay pills during playback. AI Chat: Ask follow-up questions about any insight or the article as a whole. Get deeper context without leaving the page. Community Voting: Upvote or downvote detections to help improve accuracy. Community feedback trains the system to reduce false positives over time. Share & Export: Share individual insights or full analysis reports via link. Export article analyses as professionally formatted PDF reports. Author Profiling: See patterns in how specific authors use rhetorical techniques across their published work. Domain Calibration: Compare an article's framing score against the average for that publication, so you know if a piece is unusually heavy on rhetoric for its source. Dark Mode: Full dark mode support that follows your system preferences. Plans: Free: 3 scans per month. Full analysis, highlights, and Framing Score on every scan. Pro ($3.99/month or $39.99/year): 30 scans per month. Includes AI chat, PDF reports, Blind Spots analysis, and detailed explanations. Max ($8.99/month or $89.99/year): Unlimited scans with deeper analysis powered by a more advanced AI model. Includes everything in Pro plus the Rescan button, 3x faster analysis, and priority processing. Privacy: Parallax sends article text to our servers for AI analysis — this is required for the rhetorical technique detection to work. We do not store article content after analysis. Anonymous usage data helps improve detection accuracy. Email is collected only if you sign in. Full privacy policy: https://parallaxreader.com/privacy Why Parallax? Every article you read has been crafted to make you feel something. Selective framing, loaded language, and omitted context shape your worldview without you noticing. Parallax makes the invisible visible — so you can engage with ideas on your own terms, not the author's. Think of it as a spellchecker for rhetoric.