Which AI detectors claim to detect ChatGPT? The generators field of all 38 cards
23 of the 72 AI detectors in the index name ChatGPT on the vendor page recorded in the card's generators field, read 8 and 9 November 2026; 13 name GPT-4 or a GPT-4 variant and 15 name GPT-5 or a GPT-5 variant, and 12 name none of the three, because they list a model family without the product name (Pangram lists "OpenAI" and "GPT 5.6", Originality.ai "GPT-6 Astra"), detect images, video or audio rather than text, or publish no generator list at all. A name on a vendor page is a claim of coverage, not a result: the only indexed studies that ran detectors on ChatGPT output are Weber-Wulff 2023, Chaka 2023, Orenstrakh 2023 and Walters 2023 (GPT-3.5 and GPT-4 papers), and their detection rates for the same tools run from 0 of 10 to 36 of 36 (studies index).
The generators field quotes the models and products each vendor says its detector covers, from the detector page, FAQ or documentation, in the vendor's own words and order; the table below marks whether ChatGPT, GPT-4 and GPT-5 appear by name (a spelling such as "Chat GPT" or "GPT 4" counts, a variant such as GPT-4o or GPT-5.4 counts, "OpenAI" or "GPT family" alone does not) and quotes the whole list beside the source and the date read. Three cards are marked No on the index's reading of their own wording and the reason is given in the row. This page is one of the questions answered from the index; the full cards are on the detectors hub, and the claim vs evidence gap ranking sets each vendor's accuracy claim beside the studies that tested it.
| Detector | ChatGPT named | GPT-4 named | GPT-5 named | Generators field, as the vendor states it | Source | Read date |
|---|---|---|---|---|---|---|
| QuillBot | Yes | No | Yes | GPT-6; GPT-5; ChatGPT; Gemini; Claude; Llama; other LLMs | quillbot.com/ai-content-detector | 8 November 2026 |
| Grammarly | Yes | No | No | ChatGPT; GPT-6; Gemini; Claude; Microsoft Copilot; Grammarly; other tools | grammarly.com/ai-detector | 8 November 2026 |
| Scribbr | Yes | Yes | No | ChatGPT; Gemini; Copilot; GPT2; GPT3; GPT3.5; GPT4 (experimental in free tool; supported in Premium) | scribbr.com/ai-detector | 8 November 2026 |
| Winston AI | Yes | No | No | ChatGPT; Claude; Google Gemini; LLaMA; many others; Image: Midjourney, DALL-E, Stable Diffusion, Nano Banana, ChatGPT Image, Adobe Firefly | gowinston.ai/ai-content-detector | 8 November 2026 |
| Pangram | No | No | Yes | OpenAI; Anthropic; Google; Meta; DeepSeek; xAI; Claude Opus 5; GPT 5.6; Gemini 3; Grok 4.3; Claude Sonnet 5; Claude Fable 5; GPT-5.4; GPT-OSS 120B; DeepSeek V4; Llama 3.3; Mistral Medium 3.5; Kimi K2.6; Qwen 3.7 Max; Gemma 4; Nemotron 3 Ultra; GLM 5.2; 26 models benchmarked | pangram.com | 8 November 2026 |
| GPTZero | Yes | Yes | Yes | Claude; ChatGPT; GPT-3; GPT-4; GPT-5; GPT-6; Gemini; Llama; Deepseek; Grok; AI services based on those models | gptzero.me/faq | 8 November 2026 |
| Copyleaks | Yes | Yes | Yes | GPT-5; GPT-4; ChatGPT; Claude; Gemini; DeepSeek; Llama (Meta); Bloom; Rytr; Jasper; Image: OpenAI (DALL-E and GPT), Google (Imagen and Gemini), Midjourney, Stability AI (Stable Diffusion), Black Forest Labs (Flux) | copyleaks.com/ai-content-detector | 8 November 2026 |
| Phrasly | Yes | No | Yes | ChatGPT (ChatGPT 4o, ChatGPT 5 and 5.1); GPT-5; Claude (all versions); Google Gemini and Bard; Meta Llama models; Perplexity; DeepSeek; Grok; Jasper, Copy.ai, and other AI writers | phrasly.ai/ai-detector | 8 November 2026 |
| Sapling | Yes | No | Yes | GPT-5; ChatGPT; OpenAI GPT family; Claude 4.5; Claude 5.5; Gemini 2.5; Qwen3; DeepSeek-V3; Kimi; Llama; Mistral | sapling.ai/ai-content-detector | 8 November 2026 |
| JustDone | Yes | Yes | Yes | ChatGPT; GPT-3.5; GPT-4; GPT-5; Claude; Gemini; Llama; DeepSeek; additional open-source and commercial LLMs | justdone.com/ai-detector | 8 November 2026 |
| Originality.ai | No | No | No | GPT-6 Astra (Sol, Luna, Terra); GPT-6.1 Sol; Claude Fable 5; Claude Opus 5.5; Gemini 3; Gemini 4 Argon; Kimi K3; DeepSeek V4 (Flash and Pro); Grok 4.7; Llama 4 Maverick and Scout; and more | originality.ai | 8 November 2026 |
| ZeroGPT | Yes | Yes | Yes | Chat GPT; GPT 3; GPT 4; GPT 5; GPT 6; Gemini; Grok; Perplexity AI; Claude; DeepSeek; LLaMa models; 35+ LLM families | zerogpt.com | 8 November 2026 |
| Proofademic | Yes | No | Yes | ChatGPT; GPT-5; Claude; Gemini; other LLMs; Research page: eight frontier LLMs from OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, Alibaba, xAI | proofademic.ai | 8 November 2026 |
| Walter Writes | Yes | Yes | Yes | ChatGPT (GPT-3, GPT-4, GPT-5); Claude (Anthropic); Gemini (Google); Llama (API testing) | walterwrites.ai/ai-detector | 8 November 2026 |
| Hive | No | No | No | Text: not published (single ai_generated class); Image and video source head: sora, pika, kling, luma, runway, flux, dalle, stablediffusion, midjourney, imagen, adobefirefly, bingimagecreator, ideogram, gemini, veo3, imagen4, sora2, gptimage1_5, grok and about 90 further classes; Music attribution: duobao, google, heartmula, loudly, minimax, mubert, mureka, musicgen, riffusion, stable_audio_open, suno, udio, yue, ace_step | docs.thehive.ai/docs/ai-image-and-video-d... | 8 November 2026 |
| Turnitin | No | Yes | Yes | Amazon Nova 2 Lite; Claude-3-Haiku; Claude Sonnet-3.5; Claude Sonnet-3.7; Claude-Haiku-4.5; Claude Opus 4.6; Claude Sonnet 4.5; Claude Sonnet 4.6; DeepSeek V3.2; Gemini-1.0-pro; Gemini-1.5-pro; Gemini-2.0-flash; Gemini-2.5-flash; Gemini 2.5 Pro; Gemini-3-flash-preview; Gemini-3-pro-preview; Gemini-3.1-pro-preview; GPT-4o; GPT-4o-mini; GPT-5; GPT-5-mini; GPT-5-nano; GPT-5.1; GPT-5.2; GPT-5.2 Pro; GPT-5.3; GPT-5.4; GPT-5.4-pro; o1-mini; Grok-4.1; Llama-3.3-70b; Llama-4-Maverick-17b; Mistral-large-3; Spanish model: GPT-3.5 and GPT-4; Japanese model: GPT-4 (0613), GPT-4o, GPT-4o-mini | guides.turnitin.com/hc/en-us/articles/284... | 8 November 2026 |
| GLTR | No | No | No | GPT-2 (the public demo analyses text with the GPT-2 117M model; the paper validates the method on GPT-2 1.5B samples at temperature 0.7 and top-40 truncation, and on Washington Post Heliograf articles); BERT is offered as an alternative backend model | gltr.io | 9 November 2026 |
| AIDetectPlus | Yes | No | No | ChatGPT; Gemini; Claude; any GPT-family model | aidetectplus.com/ai-detector | 9 November 2026 |
| GPT-2 Output Detector | No | No | No | GPT-2 only; the model card says it is "strongly suggested" not to use it as a ChatGPT detector | huggingface.co/openai-community/roberta-b... | 9 November 2026 |
| Codequiry | Yes | No | Yes | ChatGPT (GPT-5.5, GPT-5.5 Instant, GPT-5.4 Pro); GitHub Copilot; Claude (Opus 4.8, Sonnet 4.6, Haiku); Grok (Grok 4.3, Grok Code); Google Gemini (3.5 Pro, 3.5 Flash); Cursor AI; Amazon Q Developer (CodeWhisperer); Meta Llama (Llama 4 Maverick, Scout); General AI Detection model for any AI-generated code patterns | codequiry.com/detect-ai-written-code | 9 November 2026 |
| WasItAI | No | No | No | Midjourney; DALL-E; GPT Image (OpenAI); Stable Diffusion; Adobe Firefly; Flux; Imagen (Google); Nano Banana; Grok; Bing Image Creator; GAN-based tools such as StyleGAN and BigGAN | wasitai.com | 9 November 2026 |
| Reality Defender | No | No | No | Not named by generator. The vendor lists manipulation types: face swaps, generated video, altered faces, synthetic speech on the audio track, and AI-agent callers | realitydefender.com/deepfake-video-detection | 9 November 2026 |
| Writer | Yes | Yes | No | ChatGPT; GPT-4 (page title: ChatGPT detector and AI checker for GPT-4) | web.archive.org/web/20251209115654/https:... | 9 November 2026 |
| Undetectable AI | Yes | Yes | Yes | ChatGPT; GPT-4; GPT-5; Claude; Gemini; DeepSeek; Llama; other leading AI writing models | undetectable.ai/ai-detector | 9 November 2026 |
| AI or Not | Yes | No | No | Nano Banana; Midjourney; DALL-E / GPT Image; Stable Diffusion; Flux; Grok / Aurora; Ideogram; Adobe Firefly; Recraft; Seedream; Qwen; StyleGAN and GAN models (image coverage last updated July 2026); ChatGPT, Claude and Meta LLaMa (text) | aiornot.com/ai-image-detector | 9 November 2026 |
| Grammarly Authorship | No | No | No | A process record: ChatGPT and Claude are named only as clipboard sources it attributes, not as generators it detects | support.grammarly.com/hc/en-us/articles/2... | 9 November 2026 |
| Turnitin Clarity | No | No | No | No generators entry | turnitin.com/products/feedback-studio/cla... | 8 November 2026 |
| SynthID | No | No | No | Google models: Imagen, Veo, Lyria, Gemini and NotebookLM audio (watermarked at generation); NVIDIA (Cosmos video, per the May 2025 announcement); OpenAI (partner listed on synthid.com); Kakao (partner listed on synthid.com); Apple (announced as coming soon on deepmind.google) | synthid.com | 9 November 2026 |
| Smodin | Yes | No | Yes | ChatGPT and GPT models (page names GPT-5 in its title); Claude; Gemini; Llama; Copilot (page title); Other popular AI language models (vendor wording) | smodin.io/ai-content-detector | 9 November 2026 |
| BrandWell | Yes | Yes | No | ChatGPT; GPT-4; Claude Sonnet; Google Gemini; Various other language models (vendor wording, archived page) | web.archive.org/web/20260809041054/https:... | 9 November 2026 |
| Surfer | Yes | No | No | ChatGPT; Claude; Google Gemini; Other AI content generators, including content altered by writing tools or AI rewriters (vendor wording) | surfer.ai/ai-content-detector | 9 November 2026 |
| Compilatio | Yes | No | No | ChatGPT; Gemini; Claude; Other generative AI systems (vendor wording); the v4.6.2 model was trained against 2026 versions of ChatGPT, Claude and Gemini | compilatio.net/en/whats-new/introduction-... | 9 November 2026 |
| Quetext | Yes | Yes | Yes | ChatGPT and GPT-5 (OpenAI); Claude; Gemini (Google); Llama (Meta); Mistral; Bard, GPT-3 and GPT-4 (older page copy) | quetext.com/ai-content-detector | 9 November 2026 |
| TruthScan | Yes | Yes | No | ChatGPT; GPT-4; Claude; Bard; Jasper; Copy.ai; Hundreds of other AI writing tools (vendor wording, text FAQ); Images: 500+ generators supported (pricing page), naming Nano Banana 2, Seedream 4.5, Imagen 4, Grok, Flux 2 Pro, GPT-Image 1.5 and Midjourney | truthscan.com/ai-detector | 9 November 2026 |
| DetectGPT | No | No | No | Any model whose token log probabilities can be read; the paper evaluated GPT-2, OPT, GPT-Neo, GPT-J, GPT-NeoX, GPT-3 and Jurassic-2, and the ChatGPT mention is a study's, not the authors' | arxiv.org/pdf/2301.11305 | 9 November 2026 |
| Sightengine | No | Yes | No | Nano Banana and Nano Banana Pro (Google); GPT image generation (OpenAI, GPT-4o and GPT-1.5 image); DALL-E 2 and 3; Midjourney v5, v6, v7; Stable Diffusion (SD 1.5, 2.1, SDXL, SD3); Flux (Flux.1 Dev, Schnell, Pro); Imagen 2, 3 (Google); Ideogram 2.0, 3.0; Adobe Firefly 2, 3; Grok Imagine (xAI); Seedream (ByteDance), Reve, Recraft, Krea, Qwen, HiDream, Higgsfield, Hunyuan, Kling, MAI-Image, Vidu, Z-image, Cosmos; StyleGAN2 and StyleGAN3 and other GAN face generators; Video: Sora, Veo, Kling, Runway Gen-2 to Gen-4, Pika, Hailuo, Seedance, Wan, LTX, Midjourney Video and others | sightengine.com/detect-ai-generated-images | 9 November 2026 |
| Deepware | No | No | No | AI-generated face manipulations: deepfakes that use a real video of a person and modify or swap their face with another person (FAQ). The FAQ states the scanner does not detect voice manipulation, images or text | deepware.ai/us/faq/index.html | 9 November 2026 |
| Sensity | No | No | No | Face manipulation: AI face swap, AI lip sync, AI face reenactment, AI face morphing; AI-generated content: AI avatars, synthetic GAN faces, text-to-image, text-to-video; Audio: synthetic voice and AI voice cloning; The tech page shows a Model Generator attribution (example: Flux) beside the AI-generated content verdict; no generator list is published | sensity.ai/deepfake-detection-for-video-i... | 9 November 2026 |
Which AI detectors name ChatGPT, GPT-4 and GPT-5 together?
7 cards name all three: GPTZero, Copyleaks, JustDone, ZeroGPT, Walter Writes, Undetectable AI, Quetext, per the vendor pages read 8 and 9 November 2026 and linked in the table. GPTZero's FAQ runs from GPT-3 to GPT-6 (gptzero.me, read 8 November 2026); Turnitin's FAQ names 35 models including GPT-4o, GPT-5 through GPT-5.4-pro and, for its Spanish model, GPT-3.5 and GPT-4, without the word ChatGPT (guides.turnitin.com, read 8 November 2026). Several lists reach past the models in public use: GPTZero, QuillBot, Grammarly and ZeroGPT name GPT-6, and Originality.ai names "GPT-6 Astra" and "GPT-6.1 Sol" (originality.ai, read 8 November 2026); the index quotes the lists as written.
Which AI detectors name no ChatGPT model?
12 cards. Pangram names OpenAI as a company and "GPT 5.6" and "GPT-5.4" as models, which the table counts under GPT-5 but not under ChatGPT (pangram.com, read 8 November 2026); Originality.ai names GPT-6 and later only (originality.ai, read 8 November 2026). Hive publishes no text generator list, "single ai_generated class" (docs.thehive.ai, read 8 November 2026); Reality Defender lists manipulation types rather than generators; SynthID detects watermarks from Google models and partners; WasItAI, Sightengine, Deepware and Sensity score images, video or audio and name image and video generators or face manipulations. Turnitin Clarity's card carries no generators field, because Clarity records the writing process and detects nothing itself unless the Originality add-on is licensed (Turnitin Clarity card). The GPT-2 Output Detector names ChatGPT only to warn against using it as a ChatGPT detector (huggingface.co, read 9 November 2026), DetectGPT's paper evaluates GPT-2, OPT, GPT-Neo, GPT-J, GPT-NeoX, GPT-3 and Jurassic-2 (arXiv 2301.11305, read 9 November 2026), and Grammarly Authorship names ChatGPT as a clipboard source it attributes, not a generator it detects (support.grammarly.com, read 9 November 2026).
What do the studies say about detecting ChatGPT text?
Two pages added 9 November 2026 set the claims beside the tests: the GPT-4 question lists the thirteen cards a record scored on GPT-4 family text, and the generators covered ranking counts every model each vendor names, ChatGPT among them.
Four indexed records ran detectors on ChatGPT or GPT-4 output and the results do not follow the claims. In Weber-Wulff 2023, on 18 unedited ChatGPT documents of 2,000 characters or more, Turnitin missed 0, DetectGPT and Compilatio 1 each and Content at Scale (now BrandWell) 18, per Table 10 and the per-class counts (arXiv 2306.15666, read 9 November 2026; Weber-Wulff 2023 record). In Chaka 2023, on 6 English ChatGPT responses, Copyleaks classified 5 as AI and GLTR 1 (DOI 10.37074/jalt.2023.6.2.12, read 9 November 2026; Chaka 2023 record). In Orenstrakh 2023, on 20 English ChatGPT coursework submissions, Copyleaks caught 95% and GLTR, the GPT-2 Output Detector and GPTZero cleared 10 of 10 Spanish ChatGPT theses as human (arXiv 2307.07411, read 9 November 2026; Orenstrakh 2023 record). In Walters 2023, on 42 GPT-3.5 and 42 GPT-4 papers, Copyleaks, Turnitin and Originality.ai missed 0 while Content at Scale called 15% human (DOI 10.1515/opis-2022-0158, read 9 November 2026; Walters 2023 record). The false negative rate ranking sorts those miss rates.
Questions about AI detectors and ChatGPT
Can AI detectors detect ChatGPT?
23 of 38 vendors claim to, by name, on pages read 8 and 9 November 2026. The four indexed studies that tested ChatGPT text found miss rates from 0 of 18 (Turnitin, Weber-Wulff 2023) to 18 of 18 (Content at Scale, now BrandWell, same study), so the claim and the result are separate facts on every card.
Which AI detector is best for ChatGPT?
The index ranks no detector as best. On ChatGPT text, Turnitin missed 0 of 18 unedited documents in Weber-Wulff 2023 and Copyleaks, Turnitin and Originality.ai missed 0 of 84 GPT-3.5 and GPT-4 papers in Walters 2023 (arXiv 2306.15666; DOI 10.1515/opis-2022-0158, read 9 November 2026); the same studies' false positive rates sit beside each miss rate on the false negative rate ranking.
Do AI detectors detect GPT-5?
15 of 38 cards name GPT-5 or a variant on the vendor page, read 8 and 9 November 2026 (the table marks each). No indexed study has tested any detector on GPT-5 output; the newest generators in any record are the GPT-4o and Claude 3.5 papers of Van Vlasselaer 2026 (studies index).
Does naming ChatGPT mean the detector was trained on it?
Not on the pages read. Compilatio is the one vendor to say which versions its model was trained against, "2026 versions of ChatGPT, Claude and Gemini" for v4.6.2 (compilatio.net, read 9 November 2026); the other lists state coverage without a training statement.
Why does the GPT-2 Output Detector say not to use it on ChatGPT?
Because it is a RoBERTa classifier fine-tuned on outputs of the 1.5B-parameter GPT-2 model, and its model card says it is "strongly suggested" not to use it as a ChatGPT detector (huggingface.co, read 9 November 2026). Orenstrakh 2023 ran it on ChatGPT coursework anyway and it cleared all 10 Spanish ChatGPT theses as human (arXiv 2307.07411, read 9 November 2026).
Sources
All read 8 November 2026 unless a later date is shown; the generators source of every card is linked in the table.
- GPTZero FAQ: https://gptzero.me/faq
- Turnitin AI writing detection FAQs: https://guides.turnitin.com/hc/en-us/articles/28477544839821
- Pangram: https://www.pangram.com/
- Originality.ai: https://originality.ai
- Hive AI image and video detection documentation: https://docs.thehive.ai/docs/ai-image-and-video-detection
- GPT-2 Output Detector model card: https://huggingface.co/openai-community/roberta-base-openai-detector (read 9 November 2026)
- DetectGPT paper: https://arxiv.org/pdf/2301.11305 (read 9 November 2026)
- Grammarly Authorship support article: https://support.grammarly.com/hc/en-us/articles/29548735595405 (read 9 November 2026)
- Compilatio AI detector v4.6.2 announcement: https://www.compilatio.net/en/whats-new/introduction-to-the-compilatio-ai-detector-v4-6-2/ (read 9 November 2026)
- Weber-Wulff et al. 2023: https://arxiv.org/pdf/2306.15666
- Chaka 2023: https://doi.org/10.37074/jalt.2023.6.2.12 (read 9 November 2026)
- Orenstrakh et al. 2023: https://arxiv.org/abs/2307.07411 (read 9 November 2026)
- Walters 2023: https://doi.org/10.1515/opis-2022-0158
Revision log: 2026-11-09: page created from the generators field of 38 cards at index version 2026.11.5; 2026-11-09: GPT-4 question and generators covered ranking linked (seventh wave)
Related pages: What AI detector does College Board use? What the AP and SAT policy pages state, Which AI detectors support Spanish? The cards that name it and the one study that tested it, AI detectors ranked by the generator models the vendor names and Questions about AI detectors, answered from the index.