An AI detection and content-originality suite launched in December 2022, aimed primarily at publishers, content marketers, and academic institutions. Beyond AI detection across ChatGPT, GPT-5, Claude, Gemini, and Grok, it bundles plagiarism checking with matched source URLs, real-time fact-checking that labels claims as likely true or likely false with reasoning, paraphrase detection for tools like Quillbot, and a Chrome-based writing replay for academic integrity workflows. API and MCP access are available on the Enterprise tier. Pricing is credit-based: a free tier allows three scans per day (up to 2,000 words each), a base subscription starts around $14.95/month, and a one-time 3,000-credit pack costs $30; Enterprise is roughly $136.58/month with API included. 1 credit is roughly 100 words. Independent testing cuts both ways: Epoch AI’s July 2026 study found a 3.8% false-positive rate on human text (Pangram and GPTZero had zero in the same test), an 18% miss rate on style-imitation text, and a 29% miss rate on academic AI text—while the vendor’s own 5.0 benchmark claims 92.38% humanizer detection. Gartner reviewers praise the all-in-one report but note needing two or three rescans before passing, and some report elevated false positives.
**Pros & Cons**:
Pros:
Combines AI detection, plagiarism checking, and fact-checking in a single scan and report, which is more efficient than stitching three tools together; matched source URLs for plagiarism and reasoning for fact-check labels make the output more actionable than a bare score; writing replay via the Chrome extension adds a process-level signal for academic integrity cases; credit-based pricing with a free daily allowance supports evaluation without upfront payment; API and MCP access on Enterprise suit teams embedding detection into their own workflows; the vendor explicitly states AI scores should not be used alone for academic discipline, which is a responsible framing.
Cons: **Independent testing shows a higher false-positive rate than competitors**—Epoch AI measured 3.8% on human text, versus zero for Pangram and GPTZero in the same study, and Gartner reviewers also report elevated false positives; **results are unstable**—Gartner reviewers describe needing to rewrite and rescan two or three times before passing, so a single scan should not be treated as final; **vendor benchmarks diverge from independent tests**—the claimed 92.38% humanizer detection does not line up with Epoch’s 18% miss rate on style-imitation text, so specific accuracy figures should not be written as fact; **API requires Enterprise**, and casual users may find the subscription cost high relative to usage; academic use requires human review and should not rely on the score alone.
**Who is it for?**:Publishers and content teams vetting submitted articles for AI generation and plagiarism; academic institutions that need AI detection plus writing-process evidence, with human review built into the workflow; marketing and SEO teams checking whether content is original and factually grounded; enterprises embedding detection into their own workflows via the API.