Content Detector AI (contentdetector.ai)
Vendor ContentDetector.AI (Content Detector, Alpharetta, Georgia) Official site contentdetector.ai How this card is sourced
Updated · AI Detector Index editors
Content Detector AI classified 63% of 126 documents correctly in Walters 2023, the lowest overall share of the 16 detectors in that study, with 10% incorrect and 27% uncertain (DOI 10.1515/opis-2022-0158, Table 4, read 8 November 2026). The vendor gave its address as Content Detector, 2925 Manor Bridge Dr, Alpharetta, GA 30004, and its archived terms are governed by the laws of "ALPHARETTA, GA, 30004, United States" (Wayback capture of contentdetector.ai/terms-and-conditions, 11 May 2026, read 9 November 2026). On 9 November 2026 the domain returned a 301 to a free Mod on the Moxby marketplace, so this card reads the last live page from the Wayback Machine capture of 11 May 2026 and states where each field comes from. The card is one of the text detectors in the detector index, and how this card is sourced explains why a tool with three study records and no live page keeps its results.
What Content Detector AI detected
| Field | Value | Source |
|---|---|---|
| Content types detected | Text | web.archive.org/web/20260511112354/https:... |
| Generators covered | ChatGPT (GPT 3.5 and GPT 4 models), GPT-4, gpt-4o and gpt-o1 generated content (ChatGPT Checker line), Gemini, Claude | web.archive.org/web/20260511112354/https:... |
| Languages | English | web.archive.org/web/20260511112354/https:... |
| Input limits | Archived page (11 May 2026): paste text (above 250 words for better accuracy), word and character counters, a stat strip reading Check Counts Unlimited and an unlimited word limit; English only. The live redirect target, the Moxby AI Content Detector Mod (read 9 November 2026), reviews up to 10,000 characters per check inside the Moxby browser extension for Chrome, Safari, Edge and Firefox. No upload format is stated on either page and no API is documented. | web.archive.org/web/20260511112354/https:... |
| Free tier | Free on the archived page: a free Chat GPT Detector and ChatGPT plagiarism checker without any limitations whatsoever, with unlimited check counts and word limit (Wayback capture of 11 May 2026); the Moxby Mod the address now resolves to requires the Moxby Free plan. No monthly allowance is published | web.archive.org/web/20260511112354/https:... |
| Integrations | Moxby browser extension (Chrome, Safari, Edge, Firefox) via the AI Content Detector Mod (live redirect target), Paste-only web tool with counters for words and characters (archived page) | moxby.com/marketplace/ai-content-detector |
| Data retention | Archived Privacy Policy (Wayback capture of 11 May 2026): When you place an order through the Site, we will maintain your Order Information for our records unless and until you ask us to delete this information; the policy states nothing about submitted text. Archived Terms, Section 2: your content (not including credit card information), may be transferred unencrypted. Moxby listing FAQ (live): The submitted text is sent through the Moxby gateway to the analysis service. Do not submit confidential or sensitive text unless your use complies with your organization's policies. | web.archive.org/web/20260511112939/https:... |
| Ownership | ContentDetector.AI; the archived footer gives Content Detector, 2925 Manor Bridge Dr, Alpharetta, GA 30004 and the archived Terms and Conditions are governed by the laws of ALPHARETTA, GA, 30004, United States (Wayback Machine capture of 11 May 2026). The archived page states: It is powered by Content at Scale's AI Detector, the tool now sold as BrandWell. On 9 November 2026 contentdetector.ai returns HTTP 301 to moxby.com/marketplace/ai-content-detector, the same free AI Content Detector Mod (version 2.1.0, publisher pankil joshi) that brandwell.ai/ai-content-detector/ redirects to; the Wayback Machine index shows a 200 on 11 May 2026 and 301s on 18 August and 11 October 2026 (an earlier 301 period ran July to August 2025). The archived page also sold the WriteWell humanizer bundle. Walters 2023 tested the tool as ContentDetector.ai; Kar 2024 and the DeepSeek detectors study of 2025 as Content Detector AI. | web.archive.org/web/20260511112842/https:... |
| Independent studies | 3 records | Studies index |
The archived page described "a probability score based on the likelihood that the text content was generated by AI tools or chatbots" and said the tool "can detect Chat GPT-generated content, as well as Gemini, and Claude", adding that "This ChatGPT Checker can detect GPT-4, gpt-4o, and gpt-o1 generated content" (Wayback capture of contentdetector.ai, 11 May 2026, read 9 November 2026). The method paragraph is unusually specific for a free tool: "This detection model was developed by combining and fine-tuning the results of Open AI's GPT 2 model with the outputs of a robustly optimized BERT pretraining approach (Roberta) model." One line below the input box settles the language question: "Currently detection is available only in English." The card records English as the one supported language.
The same archived page carries a sentence that ties this card to another. "It is powered by Content at Scale's AI Detector", the tool Weber-Wulff 2023 and Walters 2023 scored under that name and which the index cards as BrandWell. Walters 2023 ran ContentDetector.ai and Content at Scale as separate detectors and printed different figures for each (63% and 71% of 126 documents correct), so the index keeps two cards and quotes the vendor's own line about the engine. The Wayback Machine index for contentdetector.ai shows 200 responses through 11 May 2026 and 301 responses on 18 August and 11 October 2026, with an earlier redirect period in July and August 2025 (web.archive.org CDX, read 9 November 2026). The redirect lands on the "AI Content Detector" Mod on moxby.com, version 2.1.0, requirements "Free plan", published under the name pankil joshi, the same listing that brandwell.ai/ai-content-detector/ resolves to (moxby.com/marketplace/ai-content-detector, read 9 November 2026). The Moxby listing reviews "up to 10,000 characters for repetitive, uniform, or robotic patterns" inside a browser extension for Chrome, Safari, Edge and Firefox, and states: "It does not determine whether a person or AI wrote the text."
Input on the archived page was paste only, with a prompt to enter "above 250 words for better accuracy", word and character counters, and a stat strip reading "Check Counts Unlimited" and an unlimited word limit (Wayback capture of 11 May 2026, read 9 November 2026). No upload format is stated on either page and no API is documented. The nearest cards are BrandWell, whose engine the archived page names and whose detector address redirects to the same Moxby Mod, and Writer, the other detector tested in Walters 2023 whose standalone tool is no longer served at its address.
Limits and plans
| Plan | Price | Limits | Checked on |
|---|---|---|---|
| Free tool (archived page) | Free | Unlimited check counts and word limit, English only; archived 11 May 2026; the address now redirects. | |
| WriteWell (limited offer, archived page) | $199/year ($499/year struck through) | Humanizer Writer, AI to Human Rewriter, Unlimited AI Detector, 40+ AI Writing Agents (archived 11 May 2026). | |
| Moxby Free plan | Free plan | AI Content Detector Mod version 2.1.0, requirements Free plan (Moxby marketplace listing, read 9 November 2026). |
The archived page sold nothing for the detector itself. It called the tool "a free Chat GPT Detector and ChatGPT plagiarism checker without any limitations whatsoever" and listed unlimited check counts and word limit (Wayback capture of 11 May 2026, read 9 November 2026). The Moxby Mod it now resolves to needs the Moxby Free plan at 10,000 characters a check (moxby.com/marketplace/ai-content-detector, read 9 November 2026). Neither states a monthly allowance, so the card has no free-words-per-month figure and sits in the second table of the free tiers ranking. The one priced offer on the archived page was the WriteWell bundle, "$199/year" with "$499/year" struck through, whose list read "Humanizer Writer", "AI to Human Rewriter", "Unlimited AI Detector" and "40+ AI Writing Agents", the same bundle the archived BrandWell page sold at the same price.
Retention was never stated for text. The archived privacy policy covers "Device Information" and "Order Information" and says of the latter "we will maintain your Order Information for our records unless and until you ask us to delete this information" (Wayback capture of contentdetector.ai/privacy-policy, 11 May 2026, read 9 November 2026). The archived terms warn that "your content (not including credit card information), may be transferred unencrypted". The Moxby listing's FAQ answers "Is my text sent to an analysis service?" with "Yes. The submitted text is sent through the Moxby gateway to the analysis service. Do not submit confidential or sensitive text unless your use complies with your organization's policies" (read 9 November 2026). The card records no retention period and the data retention ranking quotes the Moxby line in its unranked table.
What the vendor claims
No accuracy rate is published. Archived page: AI Content Detection with Max Accuracy; ContentDetector.AI detects the ChatGPT-generated content with reasonable accuracy; any detector algorithm works on probability and are not 100% accurate. The method is described as combining and fine-tuning the results of Open AI's GPT 2 model with the outputs of a robustly optimized BERT pretraining approach (Roberta) model. The live Moxby listing states no rate and says: It does not determine whether a person or AI wrote the text.
What independent studies found
- AIDI-S-0003Walters 20232023
ContentDetector.ai: 63% of 126 documents correct, 10% incorrect and 27% uncertain (Table 4), the lowest overall share of the 16 detectors; 7% of 42 GPT-4 papers correct, the lowest in the study. Tested June 25 to July 12, 2023.
- AIDI-S-0023Kar 20242024
Scored one ChatGPT 3.5 article 78.26% AI and 21.74% human (Table 1), the eighth-highest AI reading of ten free tools; after a QuillBot paraphrase the AI score read 64.58, after Grammarly 75 and after a ChatGPT paraphrase 60.87, each on one text.
- AIDI-S-0024Alshammari and Rao 20252025
Average AI score 67.27% on 49 DeepSeek-V3 answers and 59.25% after a QuillBot paraphrase, with an average human score of 55.2% on 49 human answers, the lowest of the six tools; the authors name Content Detector AI as the tool with the most false positives without printing a count (Tables III and IV, Section V-E, March 2025).
Three records in the studies index print results for the tool. Walters 2023 ran ContentDetector.ai on 126 documents between June 25 and July 12, 2023 and reports 63% correct, 10% incorrect and 27% uncertain overall, and 7% correct on the 42 GPT-4 papers, the lowest GPT-4 figure of the 16 detectors against a study average of 61% (DOI 10.1515/opis-2022-0158, Table 4, read 8 November 2026). Kar 2024 read one ChatGPT 3.5 article at 78.26% AI and 21.74% human on Content Detector AI, and after paraphrasing the same text the AI score read 64.58 (QuillBot), 75 (Grammarly) and 60.87 (ChatGPT), each on one text (DOI 10.1177/02537176241247934, Table 1, read 9 November 2026). The DeepSeek detectors study of 2025 reports an average AI score of 67.27% on 49 DeepSeek-V3 answers, 59.25% after a QuillBot paraphrase, and an average human score of 55.2% on 49 human answers, the lowest of six tools; the authors name Content Detector AI as the tool with the most false positives without printing a count (arXiv 2507.17944, Tables III and IV and Section V-E, read 9 November 2026).
None of the three figures enters a ranked field. No record prints a false positive count on human text for the tool, so the card is absent from the false positive ranking; the Walters figures are percent-correct shares whose complements the false negative ranking does not take; and the Kar 2024 and DeepSeek 2025 paraphrase figures are mean displayed scores, which feed no field under the index rule recorded with the Kar record. Pegoraro 2023 tested Content at Scale, not Content Detector AI, so that record stays on the BrandWell card.
Where Content Detector AI appears in the index
Rankings
- AIDI-R-01AI detectors ranked by independent testing Two records added in the fifth wave on 9 November 2026, Pegoraro 2023 and Fleckenstein 2024, changed nine counts and none: Pegoraro 2023 tested ZeroGPT, GPTZero, Copyleaks, Originality.ai, Writer, Content at Scale (BrandWell), DetectGPT, the GPT-2 Output Detector and GLTR, lifting GPTZero to 12, the GPT-2 Output Detector to 10, Copyleaks to 7 and Originality.ai and ZeroGPT to 6; Fleckenstein 2024 tested 289 teachers and no software, so it counts for no card ([Fleckenstein 2024 record](/studies/fleckenstein-2024-teachers/)). Illuminarty, one of four cards added in the same wave, joined with its one image record from Ha and Passananti 2024, so 18 of 72 cards now qualify.
- AIDI-R-04AI detectors ranked by languages supported
- AIDI-R-11AI detectors ranked by named integration surfaces
- AIDI-R-17AI detectors ranked by the generator models the vendor names
Fits these use cases
- Not listed in a use-case view at this index version.
Content Detector AI ranks on its three study records in independently tested detectors, appears in the GPT-4 question on the Walters 2023 figure, and on the shut-down detectors question, which records the redirect from the Wayback CDX index. On vendor-stated values read 9 November 2026 it qualifies for languages at one language for generator models named at six on the archived page, and for platform integrations at one surface, the Moxby browser extension the address now redirects to; its archived "unlimited word limit" places it in the unlimited table of the input limits ranking. It is absent from the file-type, API, LMS and retention rankings, which need a figure neither page states.
Every detector card on the register shows the same fourteen fields, so a tool that was tested three times and has since redirected can be read in the same columns as the detectors still selling at their own address.
Questions about Content Detector AI
Is Content Detector AI still available?
Not at its own address: contentdetector.ai returned HTTP 301 to the AI Content Detector Mod on moxby.com on 9 November 2026, with the Wayback Machine's last 200 response on 11 May 2026 and 301s on 18 August and 11 October 2026.
Was Content Detector AI free?
Yes. The archived page of 11 May 2026 called it "a free Chat GPT Detector and ChatGPT plagiarism checker without any limitations whatsoever", with unlimited check counts and word limit and English only.
How did Content Detector AI score in Walters 2023?
63% of 126 documents correct, 10% incorrect and 27% uncertain, the lowest overall share of 16 detectors, and 7% correct on 42 GPT-4 papers, per Walters 2023 Table 4 (DOI 10.1515/opis-2022-0158), read 8 November 2026.
Is Content Detector AI the same as Content at Scale?
The archived page said the tool "is powered by Content at Scale's AI Detector", and both addresses now redirect to the same Moxby Mod; Walters 2023 scored the two as separate detectors at 63% and 71% of 126 documents, so the index keeps two cards.
Did Content Detector AI publish an accuracy rate?
No. The archived page claimed "Max Accuracy" and "reasonable accuracy" with no percentage and said detectors "are not 100% accurate", read 9 November 2026.
Sources and revisions
Sources
- contentdetector.aiRead 8 November 2026
- web.archive.org/web/20260511112354/https:...Read 9 November 2026
- web.archive.org/web/20260511112912/https:...Read 9 November 2026
- web.archive.org/web/20260511112939/https:...Read 9 November 2026
- web.archive.org/web/20260511112842/https:...Read 9 November 2026
- web.archive.org/cdx/search/cdxRead 9 November 2026
- moxby.com/marketplace/ai-content-detectorRead 9 November 2026
- degruyterbrill.com/document/doi/10.1515/o...Read 8 November 2026
- doi.org/10.1177/02537176241247934Read 9 November 2026
- arxiv.org/abs/2507.17944Read 9 November 2026
). Illuminarty, one of four cards added in the same wave, joined with its one image record from Ha and Passananti 2024, so 18 of 72 cards now qualify.: register plate](/img/independently-tested-480.webp)
