ZeroGPT AI detector
Vendor Olive Works LLC, doing business as ZeroGPT Official site zerogpt.com How this card is sourced
Updated · AI Detector Index editors
ZeroGPT's free plan allows 15,000 characters per AI detection and 1,250,000 words per month, according to the ZeroGPT pricing page read 8 November 2026. The vendor is Olive Works LLC, doing business as ZeroGPT and organized in Wyoming, per the ZeroGPT privacy policy read the same day. To compare ZeroGPT with the other twenty-one cards, start from the full table. How this card is sourced.
What ZeroGPT detects
| Field | Value | Source |
|---|---|---|
| Content types detected | Text, image, video | zerogpt.com |
| Generators covered | Chat GPT, GPT 3, GPT 4, GPT 5, GPT 6, Gemini, Grok, Perplexity AI, Claude, DeepSeek, LLaMa models, 35+ LLM families | zerogpt.com |
| Input limits | Maximum characters per AI detection: FREE 15,000, PRO 100,000, PLUS 100,000, MAX 150,000, EDU 250,000, EXPERT 350,000, ENTERPRISE unlimited. Batch files: FREE 0, PRO 50, PLUS 60, MAX 75, EDU 80, EXPERT 100. Paste text or upload .txt, .docx, text-based .pdf. Guidance: at least 150 to 200 words. API plans: 2MB (Beginner), 5MB (PRO), 15MB (VIP) max file size. | zerogpt.com/pricing |
| Free tier | $0 forever, no credit card: 15,000 characters per AI detection, words allowance 1,250,000 /month, 0 batch files, no PDF reports, no detection history; plagiarism checker 300 words one-time-only | zerogpt.com/pricing |
| API | Yes, app.theneo.io/olive-works-llc/zerogpt-doc... | app.theneo.io/olive-works-llc/zerogpt-doc... |
| Integrations | LMS or enterprise systems via API (no named LMS), WhatsApp bot, Telegram bot | zerogpt.com/faq |
| Data retention | Privacy policy (Last Updated October 07, 2026): ZeroGPT does not use private User Content submitted through the Services to train, retrain, fine-tune, or otherwise develop AI models. Private User Content not intentionally saved as part of an account feature is retained only for the period reasonably necessary to provide and operate the applicable Service, including reasonable temporary processing, security, abuse-prevention, diagnostic, and backup requirements. Infrastructure hosted in Germany under GDPR; no-retention modes available for API/enterprise. Paid plans include History of detections. | zerogpt.com/privacy-policy |
| Ownership | Olive Works LLC, doing business as ZeroGPT, organized in Wyoming, United States. Parent company not published. | zerogpt.com/privacy-policy |
| Independent studies | 7 records | Studies index |
ZeroGPT covers text, image and video, per the ZeroGPT home page read 8 November 2026. For text, the vendor names Chat GPT, GPT 3, GPT 4, GPT 5, GPT 6, Gemini, Grok, Perplexity AI, Claude, DeepSeek and LLaMa models, and says 35+ LLM families are covered.
The ZeroGPT FAQ says paraphrasing can reduce some signals, and the terms of use say ZeroGPT does not guarantee that text processed by a humanization or rewriting tool will be classified as human-written. ZeroGPT also sells its own AI Humanizer (ZeroGPT FAQ, read 8 November 2026).
The privacy policy, last updated 7 October 2026, says ZeroGPT does not use private User Content to train models, and that content not saved as part of an account feature "is retained only for the period reasonably necessary to provide and operate the applicable Service" (privacy policy, read 8 November 2026). The FAQ says text is discarded after the result, while paid plans advertise a history of detections; the card leads with the privacy policy. Integrations are a WhatsApp bot, a Telegram bot and LMS or enterprise systems via the API. Two related cards: GPTZero and Winston AI, both also scored in Weber-Wulff 2023 and RAID.
Limits and plans
| Plan | Price | Limits | Checked on |
|---|---|---|---|
| PRO (monthly) | C$22.99 /month | 100,000 characters per detection, 50 batch files, PDF reports, history, 4,500,000 words/month, plagiarism 750 words one-time-only. | |
| PRO (annual) | C$13.99 /month billed annually = C$167.88 /year | Same as PRO (monthly). Banner: 40% off annually. | |
| PLUS (monthly) | C$32.99 /month | 100,000 characters, 60 batch files, 5,000,000 words/month, plagiarism 35,000 words/month (100 detections/month). | |
| PLUS (annual) | C$20.99 /month billed annually = C$251.88 /year | Same as PLUS (monthly). | |
| MAX (monthly) | C$46.99 /month | 150,000 characters, 75 batch files, 7,500,000 words/month, plagiarism 60,000 words/month (150 detections). | |
| MAX (annual) | C$29.99 /month billed annually = C$359.88 /year | Same as MAX (monthly). | |
| EDU | C$32.99 /month, Billed yearly only, C$395.88 /year | 250,000 characters, 80 batch files, 70,000,000 words/year, plagiarism 1,200,000 words/year (4500/year). | |
| EXPERT (monthly) | C$90.99 /month | 350,000 characters, 100 batch files, 15,000,000 words/month, plagiarism 150,000 words/month (400 detections). | |
| EXPERT (annual) | C$57.99 /month billed annually = C$695.88 /year | Same as EXPERT (monthly). | |
| ENTERPRISE | Get a quote. | ||
| API Pay as you go | Beginner AI Detection $0.042 / 1000 Words; PRO $0.057 / 1000 Words; VIP $0.072 / 1000 Words; AI Image Detection $0.013 per image | Beginner 50,000 chars per detection, 40 batch files, 2MB; PRO 150,000 chars, 75 files, 5MB; VIP 500,000 chars, 150 files, 15MB. |
Characters per detection run FREE 15,000, PRO and PLUS 100,000, MAX 150,000, EDU 250,000, EXPERT 350,000 and ENTERPRISE unlimited; uploads accept .txt, .docx and text-based .pdf; ZeroGPT suggests at least 150 to 200 words (pricing page, read 8 November 2026). Personal prices are in Canadian dollars exactly as the pricing page displayed them to a Canadian viewer on 8 November 2026; the page has no currency switcher and shows no USD personal prices. The API cards on the same page are in US dollars.
The free tier is $0 forever with no credit card: 15,000 characters per detection, 1,250,000 words per month, 0 batch files, no PDF reports or history (pricing page, read 8 November 2026).
What the vendor claims
98.4% Detection accuracy; 35+ LLM families covered (ZeroGPT product metrics, DeepAnalyse page and home). FAQ: pushing toward >98% on internal evaluations.
ZeroGPT says: "98.4% Detection accuracy" and "<1% False positive rate", labelled product metrics on the ZeroGPT DeepAnalyse page read 8 November 2026. The FAQ at zerogpt.com/faq says ZeroGPT is "pushing toward >98% on internal evaluations", and the home page FAQ says results "converge to an error rate lower than 2%". The card uses the DeepAnalyse figures because that page is the vendor's methodology-style page, and shows the FAQ figures beside them; none of the four carries a dataset or a date.
ZeroGPT's terms say: "You must not treat a ZeroGPT Output as the sole basis for a decision that could materially affect another person's rights, education, employment, admission" (terms of use, read 8 November 2026).
What independent studies found
- AIDI-S-0001Weber-Wulff 20232023
ZeroGPT: 32 of 54 documents correct (59% binary accuracy, rank 6 of 14); 72% logarithmic accuracy (rank 5); 0 false accusations in 18 human-written documents; 13 of 36 AI documents undetected (36.1%). Tested March 7 to 28, 2023.
- AIDI-S-0002Liang 20232023
ZeroGPT was one of seven detectors tested (accessed March 15, 2023). Across the seven, the paper reports an average false positive rate of 61.22% on 91 human-written TOEFL essays and near-perfect accuracy on 88 US 8th-grade essays. Per-detector rates appear only in a bar chart (Fig. 1a) and are not stated as numbers.
- AIDI-S-0003Walters 20232023
ZeroGPT: 87% of 126 documents correct, 1% incorrect, 12% uncertain; 100% correct on 42 GPT-3.5 papers, 83% on 42 GPT-4 papers; 79% of 42 human papers correct, 2% incorrect (false positive rate 2%). Tested June 25 to July 12, 2023.
- AIDI-S-0005Perkins 20242024
ZeroGPT: baseline mean accuracy 46.1% (rank 5 of 7, Table 7); 31.3% on non-manipulated AI text falling to 17.3% after adversarial techniques (drop 14%, Table 8); 0 false accusations in 10 human samples; 77 of 104 AI samples undetected (UDC 73%).
- AIDI-S-0006RAID 20242024
ZeroGPT: false positive rate 1.71% / 1.42% / 1.21% / 0.90% at naive thresholds 0.25 / 0.5 / 0.75 / 0.95 on human-written documents (Table 4); could not reach the 5% FPR target in some settings (asterisk); accuracy at 5% FPR 95.4% on open-source chat models (greedy), 0.3% on open-source non-chat with sampling and repetition penalty (Table 5).
- AIDI-S-0018Pegoraro 20232023
True positive rate 45.7% on ChatGPT responses and true negative rate 92.2% on human responses, the highest TPR of the 12 online tools in a 20-method sweep on a ChatGPT and human Q&A benchmark (Table I).
- AIDI-S-0023Kar 20242024
Read one 500-word ChatGPT 3.5 article as 95.03% AI and the same text after a QuillBot paraphrase as 32.31 AI; 96.74 after Grammarly and 90.61 after ChatGPT paraphrases (Table 1, single text).
Weber-Wulff 2023: 32 of 54 documents correct (59%, rank 6 of 14), 0 false accusations in 18 human-written documents, 13 of 36 AI documents undetected, tested 7 to 28 March 2023 (PDF, read 8 November 2026).
Liang 2023 included ZeroGPT among seven detectors accessed 15 March 2023; across the seven, the average false positive rate was 61.22% on 91 human-written TOEFL essays, with per-detector rates shown only in a bar chart (PDF, read 8 November 2026).
Walters 2023: 87% of 126 documents correct, 1% incorrect, 12% uncertain; 83% on 42 GPT-4 papers; 2% of 42 human papers labelled AI, tested June 25 to July 12, 2023 (PDF, read 8 November 2026).
Perkins 2024: baseline mean accuracy 46.1% (rank 5 of 7), falling from 31.3% to 17.3% after adversarial techniques; 0 false accusations in 10 human samples; 77 of 104 AI samples undetected (PDF, read 8 November 2026).
RAID benchmark (2024): false positive rate 1.71%, 1.42%, 1.21% and 0.90% at thresholds 0.25, 0.5, 0.75 and 0.95; accuracy at a fixed 5% FPR 95.4% on open-source chat models with greedy decoding and 0.3% on non-chat models with sampling and a repetition penalty (PDF, read 8 November 2026).
Zero false accusations in Weber-Wulff 2023 and Perkins 2024, 2% in Walters 2023, 1.42% in RAID at threshold 0.5: four studies, four bases, four answers. The base is as much a part of the result as the rate.
Pegoraro 2023, added 9 November 2026, ran 20 detection methods on a ChatGPT and human question-and-answer benchmark in March 2023 and records ZeroGPT at a 45.7% true positive rate on ChatGPT responses and a 92.2% true negative rate on human responses, the highest true positive rate of the 12 online tools in the sweep (arXiv 2304.01487, Table I, read 9 November 2026); the rates are printed as rates, so the ranking fields take no value from them.
Kar 2024 joined the list on 9 November 2026 and took the count to 7: one 500-word ChatGPT 3.5 article read at 95.03% AI, then 32.31 after a QuillBot paraphrase, 96.74 after Grammarly and 90.61 after ChatGPT, single-text scores that move no rank field.
Where ZeroGPT 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-03AI detector free tiers, compared
- AIDI-R-07AI image, voice, video and code detectors ranked by independent studies
- AIDI-R-08AI detectors ranked by published robustness to paraphrased and edited text
- 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-14AI detector APIs ranked by published price per 1,000 words
- 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-03AI detectors for copywriters and agencies
- 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-09AI detectors for law firms and courts: an index fit check
- AIDI-U-10AI detectors for grant and proposal reviewers
- AIDI-U-13AI detectors for government and public sector buyers
- AIDI-Q-01Is there a 100% accurate AI detector? What published tests show
- 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-03ZeroGPT vs Turnitin: the 14 fields and the three shared studies side by side
- AIDI-C-07ZeroGPT vs Scribbr vs QuillBot: three free AI detectors on the same 14 fields
ZeroGPT appears in three rankings: AI detectors ranked by independent testing with five studies, AI detector false positive rates with 0 false accusations in 18 human documents from Weber-Wulff 2023, and AI detector free tiers with 1,250,000 words per month from the pricing page read 8 November 2026.
ZeroGPT sits on AI detectors for business and AI detectors for copywriters and agencies because of the API, and on AI detectors for job applications because of the free tier. The detector hub lists all twenty-two cards. A reader who wants a pay-per-word alternative to the free monthly allowance can compare AIDetectPlus, which sells 5,000 credits for $9.00 one time, per its home page read 9 November 2026.
Questions about ZeroGPT AI detector
Is ZeroGPT free to use?
Yes, with limits. ZeroGPT's free plan is $0 forever with no credit card and allows 15,000 characters per detection and 1,250,000 words per month, with no batch files, PDF reports or history, per the pricing page read 8 November 2026.
Why are ZeroGPT's prices shown in Canadian dollars?
Because the pricing page geo-prices and showed Canadian dollars to the viewer who read the page on 8 November 2026, with no currency switcher. The card repeats the prices exactly as displayed, from C$22.99 per month for PRO to C$90.99 for EXPERT.
What false positive rate did ZeroGPT show in studies?
Zero false accusations in 18 human documents in Weber-Wulff 2023 and in 10 in Perkins 2024, 2% of 42 human papers in Walters 2023, and 1.42% at threshold 0.5 in RAID. ZeroGPT itself says "<1% False positive rate" on the DeepAnalyse page read 8 November 2026.
How did ZeroGPT perform on the RAID benchmark?
At a fixed 5% false positive rate, ZeroGPT scored 95.4% on open-source chat models with greedy decoding and 0.3% on non-chat models with sampling and a repetition penalty, per RAID (2024). RAID notes ZeroGPT could not reach the 5% FPR target in some settings.
Sources and revisions
Sources
- zerogpt.comRead 8 November 2026
- zerogpt.com/pricingRead 8 November 2026
- zerogpt.com/privacy-policyRead 8 November 2026
- zerogpt.com/faqRead 8 November 2026
- zerogpt.com/deep-analyseRead 8 November 2026
- zerogpt.com/terms-of-useRead 8 November 2026
- arxiv.org/pdf/2306.15666Read 8 November 2026
- arxiv.org/pdf/2304.02819v3Read 8 November 2026
- degruyterbrill.com/document/doi/10.1515/o...Read 8 November 2026
- arxiv.org/pdf/2403.19148Read 8 November 2026
- arxiv.org/pdf/2405.07940Read 8 November 2026
Revision log
- Card created2026-11-09: pegoraro 2023 record added to the study list in the fifth wave2026-11-09: kar 2024 record added in the sixth wave, 7 study records
). 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)
