Winston AI detector
Vendor Winston AI inc. Official site gowinston.ai/ai-content-detector How this card is sourced
Updated · AI Detector Index editors
Winston AI detects AI text in 11 languages, plus AI images and handwriting through OCR, according to the detector page read on 8 November 2026 (gowinston.ai). Two study records in the index cover Winston AI. How this card is sourced explains the rules; the full detector index holds the other twenty-one cards.
What Winston AI detects
| Field | Value | Source |
|---|---|---|
| Content types detected | Text, image, handwriting | gowinston.ai/pricing |
| Generators covered | ChatGPT, Claude, Google Gemini, LLaMA, many others, Image: Midjourney, DALL-E, Stable Diffusion, Nano Banana, ChatGPT Image, Adobe Firefly | gowinston.ai/ai-content-detector |
| Languages | English, French, Spanish, Portuguese, German, Dutch, Polish, Italian, Indonesian, Romanian, Chinese (Simplified) | gowinston.ai/ai-content-detector |
| Input limits | Characters per scan: up to 200,000 (Essential, Advanced, Elite; Free row blank). Scans per hour: unlimited on paid plans. Paste text or upload documents in .docx, .png and .jpg. File size not published. | gowinston.ai/pricing |
| Free tier | Free: $0 /month, 2,000 credits / 14 day trial (not recurring). API: register for free and receive 2,500 credits. | gowinston.ai/pricing |
| API | Yes, docs.gowinston.ai/api-reference/introduction | docs.gowinston.ai/api-reference/introduction |
| Integrations | Chrome Extension, Google Classroom, WordPress, Zapier, API, GPT apps in ChatGPT | gowinston.ai/ai-content-detector |
| Data retention | Privacy Policy (Last updated July 17, 2026): We collect and store any text or images you submit for analysis through any of our features. Text submitted for analysis is not shared with or compared against other users' submissions. If you delete a report, the associated submitted content is deleted from active storage. If you delete your account, all associated submitted content and reports are deleted. We do not use your individually identifiable submitted content to train AI models without your explicit consent. | gowinston.ai/privacy-policy |
| Ownership | Winston AI inc.; no parent company stated | gowinston.ai/privacy-policy |
| Independent studies | 2 records | Studies index |
Winston AI covers text, images and handwriting (OCR). For text the detector page names ChatGPT, Claude, Google Gemini and LLaMA; for images, Midjourney, DALL-E, Stable Diffusion, Nano Banana, ChatGPT Image and Adobe Firefly (gowinston.ai/ai-content-detector/, read 8 November 2026).
The 11 languages in the detector page FAQ are English, French, Spanish, Portuguese, German, Dutch, Polish, Italian, Indonesian, Romanian and Chinese (Simplified) (gowinston.ai/ai-content-detector/, read 8 November 2026). The page body lists 10, the home page adds Tagalog, Russian and Bulgarian, and the pricing table lists Ukrainian; this card uses the FAQ list, the longest on the product page.
On edited AI text, Winston AI says: "Identify AI-generated writing even when its structure and wording have been substantially modified" (gowinston.ai/ai-content-detector/, read 8 November 2026).
Winston AI inc. owns the product. Integrations are a Chrome extension, Google Classroom, WordPress, Zapier, GPT apps in ChatGPT and an API documented at docs.gowinston.ai (gowinston.ai/ai-content-detector/, read 8 November 2026).
Limits and plans
| Plan | Price | Limits | Checked on |
|---|---|---|---|
| Essential (monthly) | $18 $9 /month | List price struck through, promo price with code WINSTON50 (50% off). 100,000 credits / month; AI content detection, AI image and deepfake detection, plagiarism detection, writing feedback, OCR, PDF reports. | |
| Advanced (monthly) | $29 $14.50 /month | 200,000 credits / month; up to 5 team members; all Essential features. | |
| Elite (monthly) | $49 $24.50 /month | 500,000 credits / month; unlimited team members; all Essential features. | |
| Essential (annual) | $10.00 $5.00 /month, $60.00/year | Toggle reads Annual Save 45%. 100,000 credits / month. | |
| Advanced (annual) | $16.00 $8.00 /month, $96.00/year | 200,000 credits / month; up to 5 team members. | |
| Elite (annual) | $26.00 $13.00 /month, $156.00/year | 500,000 credits / month; unlimited team members. | |
| Enterprise | Contact sales. |
Paid plans take up to 200,000 characters per scan with unlimited scans per hour; the Free row of the pricing table is blank for characters per scan. Input is pasted text or .docx, .png and .jpg files (Winston AI pricing page, read 8 November 2026).
The free tier is $0 /month with 2,000 credits for a 14 day trial, granted once; API sign-up adds 2,500 free credits (gowinston.ai/pricing/, read 8 November 2026).
Prices below are as displayed on the pricing page on 8 November 2026: a struck-through list price beside a promotional price tied to the code WINSTON50, with a dollar symbol and no currency named.
On retention, the privacy policy of July 17, 2026 says: "We collect and store any text or images you submit for analysis", that deleting a report deletes the content from active storage, and that identifiable content is not used to train models "without your explicit consent" (privacy policy, read 8 November 2026). Annual plans cannot be cancelled during the term (gowinston.ai/ai-content-detector/, read 8 November 2026).
What the vendor claims
Winston ranks first on RAID, achieved 100% accuracy on the Chicago Booth benchmark, and recorded zero false positives across 27,743 human documents in published evaluations. 99.89% of AI text detected at a 1% false-positive rate, including adversarial attacks. (Headline elsewhere on the home page: The only AI detector with a 99,87% accuracy rate.)
Winston AI inc. says: "Winston ranks first on RAID, achieved 100% accuracy on the Chicago Booth benchmark, and recorded zero false positives across 27,743 human documents in published evaluations. 99.89% of AI text detected at a 1% false-positive rate, including adversarial attacks." The sentences are from the Winston AI home page, read 8 November 2026. A headline on the same page reads "The only AI detector with a 99,87% accuracy rate"; this card quotes the research block, which names the benchmark.
On false positives, Winston AI inc. says: "Recorded zero false positives across 27,743 human documents in published evaluations. PERSUADE 2.0: 25,996 genuine student essays, zero false positives. Tuned to eliminate false positives." (gowinston.ai, read 8 November 2026). The same page gives a 1% false-positive operating point for the 99.89% detection figure; this card carries both and ranks Winston AI on study data only.
Winston AI inc. also says: "AI detectors have been developed to determine if online content is human or machine-generated, though their effectiveness currently varies" (gowinston.ai/ai-content-detector/, read 8 November 2026). Zero false positives on a benchmark is a statement about that benchmark, not about the next document a reader scans.
What independent studies found
- AIDI-S-0001Weber-Wulff 20232023
Go Winston: 36 of 54 documents correct (67% binary accuracy, rank 5 of 14); 75% logarithmic accuracy (rank 4); 0 false accusations in 18 human-written documents; 12 of 36 AI documents undetected (33.3%), mostly manually edited or Quillbot-paraphrased texts. Tested March 7 to 28, 2023.
- AIDI-S-0006RAID 20242024
Winston: false positive rate 0.75% / 0.55% / 0.38% / 0.21% at naive thresholds 0.25 / 0.5 / 0.75 / 0.95 on human-written documents (Table 4); accuracy at a fixed 5% FPR 97.2% on open-source chat models (greedy decoding), 68.2% on open-source non-chat models (greedy), 29.5% on open-source non-chat with sampling (Table 5).
Weber-Wulff 2023 scored Go Winston at 36 of 54 documents correct (67%, rank 5 of 14), with 0 false accusations in 18 human-written documents and 12 of 36 AI documents undetected, mostly manually edited or Quillbot-paraphrased texts, in testing from March 7 to 28, 2023 (arXiv 2306.15666, read 8 November 2026).
The RAID benchmark (2024) reports a Winston false positive rate of 0.75%, 0.55%, 0.38% and 0.21% at thresholds 0.25, 0.5, 0.75 and 0.95 on human-written documents, and accuracy at a fixed 5% FPR of 97.2% on open-source chat models with greedy decoding, 68.2% on non-chat models with greedy decoding, and 29.5% on non-chat models with sampling (arXiv 2405.07940, read 8 November 2026).
Where Winston 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-02AI detector false positive rates, ranked
- AIDI-R-04AI detectors ranked by languages supported
- AIDI-R-05AI detectors ranked by maximum input per check
- AIDI-R-07AI image, voice, video and code detectors ranked by independent studies
- AIDI-R-09AI detector false negative rates, ranked
- AIDI-R-10AI detector accuracy claims vs published results, ranked by the gap
- AIDI-R-11AI detectors ranked by named integration surfaces
- AIDI-R-12AI detectors ranked by the upload formats the vendor states
- AIDI-R-15AI detectors ranked by the LMS platforms the vendor names
- AIDI-R-17AI detectors ranked by the generator models the vendor names
- AIDI-R-18AI detectors ranked by study records on non-English and non-native text
Fits these use cases, questions and comparisons
- AIDI-U-01AI detectors for business and managers
- AIDI-U-02AI detectors for school districts
- AIDI-U-04AI detectors for job applications and recruiters
- AIDI-U-06AI detectors for admissions offices
- AIDI-U-08AI detectors for researchers checking their own manuscripts
- AIDI-U-10AI detectors for grant and proposal reviewers
- AIDI-U-12AI detectors for localisation and translation teams
- AIDI-Q-31Which AI detectors were in the Weber-Wulff 2023 study? The 14 tools, their result on 54 documents, and what each one is today
- AIDI-C-05Winston AI vs Turnitin: the 14 fields and the study results side by side
- AIDI-C-09Winston AI vs Originality.ai: the 14 fields and the study results side by side
Winston AI qualifies for three rankings: independently tested on 2 studies, false positive rate on the Weber-Wulff 2023 figure of 0 in 18, and languages on a count of 11 (detectors.json, read 8 November 2026). The free tier is a one-time trial, so the card is absent from the free tiers ranking.
Winston AI fits business through the API and team seats, school districts through Google Classroom, agencies through the API, Zapier and WordPress, and recruiters through the free trial on plain text (gowinston.ai/ai-content-detector/, read 8 November 2026). Other image-capable cards are Pangram and Copyleaks. The home page carries every AI detector we track.
Two comparisons added 9 November 2026 read this card against other tools on their one shared record each: Winston AI vs Turnitin on Weber-Wulff 2023 and Winston AI vs Originality.ai on RAID 2024.
Questions about Winston AI detector
How much does Winston AI cost?
$18 /month list, shown as $9 /month with the promotional code, for Essential billed monthly, per the pricing page read 8 November 2026. Annual Essential shows $10.00 struck to $5.00 /month, $60.00/year. Advanced and Elite list at $29 and $49 monthly.
Is Winston AI free?
Only as a trial. The free tier is 2,000 credits for 14 days at $0 /month, with no recurring allowance, per the pricing page read 8 November 2026. API registration adds 2,500 credits.
What false positive rate do studies give Winston AI?
0 of 18 human-written documents in Weber-Wulff 2023, and 0.55% at threshold 0.5 in RAID 2024, read 8 November 2026. The vendor's own claim is zero across 27,743 human documents; the card ranks on the Weber-Wulff count.
Sources and revisions
Sources
- gowinston.ai/ai-content-detectorRead 8 November 2026
- gowinston.aiRead 8 November 2026
- gowinston.ai/pricingRead 8 November 2026
- gowinston.ai/ai-content-detection-apiRead 8 November 2026
- gowinston.ai/privacy-policyRead 8 November 2026
- docs.gowinston.ai/api-reference/introductionRead 8 November 2026
- arxiv.org/pdf/2306.15666Read 8 November 2026
- arxiv.org/pdf/2405.07940Read 8 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)
