Four commitments
that shape every scan
Forensic, Not Editorial
We score what is structurally present — framing, omissions, emotional load — not whether we agree with the conclusion.
Multi-Axis, Not Left/Right
A single label flattens complexity. We decompose text into independent rhetorical dimensions (Aristotelian appeals, 24 fallacies).
Span-Anchored Evidence
Every score traces back to exact words. "Why this 72?" is always answerable with a char-offset quoted passage.
Reproducible & Auditable
Deterministic aggregation with zero LLM hallucination. Every scan returns a prompt_hash and fme_version.
While others check facts,
we check the framework
Bias Spectrum Mapping
Visualize any article's political lean on a precision gradient from Far Left to Far Right. See exactly where content lands — beyond a simple label.
Propaganda Index
Quantifies emotional loading, logical fallacy density, and intent transparency in a single score.
Strategic Silence
Surfaces omitted perspectives, missing counter-evidence, and deliberately absent context.
Fallacy Density Index
Detects 24 types of logical fallacies with exact quoted text and impact analysis.
The Unified
Single-Call Pipeline
Article Ingestion
Word count gate (≥250 words), cache lookup. 80%+ cache hit rate → instant result with zero new LLM calls.
Unified LLM Call
Single gpt-4.1-nano call with V20 system prompt. One call returns ideology scores, spans, emotional resonance, and credibility signals.
Calibration Layer
9 deterministic rule-based overrides correct nano's known misclassifications — biometric privacy → Libertarianism, wire news → Technocratic Governance, and 7 more.
Validated Output
Zod schema validation + V20 metadata injection. Delivers FME V20 analysis to extension panel, API, and 11-page PDF export.
80%+ Instant Results
V20 ships a production cache layer that returns prior analysis results in under a second. For the 80%+ of articles already scanned by the community, there's no LLM call — just instant forensic insight.
Built for anyone who reads media critically
Whether you analyze for a living or just want to read with clarity, Rhetoric Audit gives you the forensic tools to understand what you're really reading.
Pre-coded rhetorical analysis at scale
Save hundreds of hours in literature reviews. Rhetoric Audit pre-codes articles for ideological framing, logical consistency, and epistemic humility — so you can focus on synthesis, not classification.
Verify narratives, not just facts
Facts can be true while the narrative remains manipulative. Rhetoric Audit exposes the rhetorical architecture — framing, emotional loading, source reliability — so you can report with full context.
Measure media framing on policy topics
Understand how policy issues are being shaped in media before they reach public opinion. Rhetoric Audit's volatility correlation and philosophical framing reveal the ideological vectors at play.
Read with your eyes open
Every article is an argument. Rhetoric Audit gives you the tools that PhD researchers use — in your browser — so you can understand not just what an article says, but how it's trying to make you think.
From install to insight
in under a minute
Install the Extension
Add Rhetoric Audit to Chrome in one click. Free, no API key required, Manifest V3 secure.
Create Your Account
Sign up with your email and receive 10 free analysis credits instantly — no credit card needed.
Analyze Any Article
Browse to any news article. Click the Rhetoric Audit icon and receive a full forensic report in 2–5 seconds.
15 diagnostic parameters.
One forensic report.
Intelligence Brief
Cross-platform narrative intelligence in one report. Pull signals from X, Reddit, YouTube, News, and the open Web — quantify authenticity, emotion, and contagion risk in minutes.
2 credits per brief · PDF export included
Organic engagement with coordinated clusters detected
Trusted by researchers,
journalists & critical thinkers
"This tool provides a bridge between critical theory and digital media. It doesn't just tell you what the article says; it tells you how it wants you to think."
"The Strategic Silence feature alone is worth it. Knowing what a journalist chose NOT to include is often more revealing than the article itself."
"I use Rhetoric Audit in my media literacy class. Students go from passive consumers to active analysts in one session."
How We Compare
Manual analysis is slow and subjective. Other tools lack forensic depth. Rhetoric Audit combines speed, precision, and reproducibility at scale.
| Criterion | Manual Analysis | Other AI Tools | Rhetoric Audit |
|---|---|---|---|
| Time to Analysis | 4–8 hours per article | 30 sec – 2 min | 2–5 seconds (cached: <1s)✓ |
| Cost per Article | $180–240 (researcher labor) | $0.01–0.05 | $0.005 or free✓ |
| Accuracy on Bias Detection | Variable (subjective judgment) | 70–85% | 93.3% (production-verified, 21-article benchmark)✓ |
| Reproducibility | Low (different analysts may disagree) | Medium | 100% (same article = same result always)✓ |
| Evidence Trail | Notes are analyst-specific, hard to defend | — | Every score anchored to exact quoted text (char-offsets)✓ |
| Scale | 1–3 articles per day per analyst | 100s per day | 1,000s per day, $9/mo unlimited✓ |
| Detection of Propaganda Techniques | Requires training in SemEval-2020 labels | — | Automated span-level annotation (24 fallacy types)✓ |
| Strategic Silence (Omissions) | Requires domain expertise to spot absences | — | Flagged automatically with context gaps✓ |
138x ROI in Year 1
A think tank reduced media analysis from 40 hours/week to 4 hours/week using Rhetoric Audit. Annual savings: $180,050.
A newsroom tracked 6-month adoption: 37x ROI, +12% organic traffic, +18% reader trust.
Read case studies →Real Results from
Real Teams
See how researchers and journalists accelerate analysis with Rhetoric Audit — and the measurable impact on their work.
Think Tank Cuts Analysis Time 90%
A policy research team reduced media analysis from 40 hours per week to 4 hours using Rhetoric Audit — saving $180,050 annually.
Newsroom Boosts Reader Trust 18%
A newsroom tracked 6 months of Rhetoric Audit use: 37x ROI, +12% organic traffic, +18% reader trust, real-time propaganda detection.
Start reading media
with forensic clarity
Join researchers, journalists, and critical readers who use Rhetoric Audit to understand not just what an article says — but how it's engineered to make you think.