AI detectors for copywriters and agencies

AIDI-U-03

Updated ·

An AI detector for copywriters has to handle volume: dozens of drafts a week, checked before a client runs them through a detector of their own. Two fields on the index decide fit for that job: api, so checks can run inside a workflow, and a bulk or batch upload stated in input_limits, integrations or plans. Of the 72 cards, 39 state an API, 13 state a bulk or batch feature, 28 state at least one of the two, and 11 state both. Those 11 cards made up this view at version 2026.11.9 and Isgen joined on 10 November 2026 with the ninth wave (a 150-file Bulk Scan and a RapidAPI listing), so 12 qualified at that version, and Ircam Amplify joined later the same day with the tenth wave (API docs and a 20-file batch ceiling, for music), so 13 qualify; Smodin, Quetext and TruthScan joined with the fourth wave of cards on 9 November 2026, with price per word or per plan and the free tier carried as checked. This page is one of five views under AI detectors by use case; the business view filters on team seats and retention, the code view on code modality, and the card for each tool sits on the full board.

  1. Field 1Bulk or batch as stated
  2. Field 2API or plan price as checked
  3. Field 3Free tier as stated
  4. Field 4Studies

Volume and price. The API field is a yes only when the vendor publishes developer docs. The bulk or batch value is quoted as the vendor states the ceiling: files per upload, documents per CSV, or a Bulk Scan feature by name. Price is the metered API rate where one is published, and the entry plan with its monthly word allowance where none is. The free tier is the allowance the vendor states, the field the free tiers ranking orders the whole index on. Field rules are in the methodology. Undetectable AI and AI or Not, added on 9 November 2026, publish API docs but state one pasted text or one uploaded file per check and no bulk or batch ceiling (undetectable.ai, docs.aiornot.com, read 9 November 2026), so neither enters this view.

Which fields decide fit for an agency?

AI detectors for copywriters and agencies: the cards in this view and the fields that decide fit
DetectorBulk or batch as statedAPI or plan price as checkedFree tier as statedStudies
Pangramup to 100 files at a timeText (Pangram 4.1): $0.05 per 100 words20 free credits daily0
GPTZeroa batch of up to 250 files (free account)premium plans starting at $12.99/monthFree plan covering 10,000 words per month6
Phraslybulk scans from a CSV of up to 100 documents (Business API)$0.02 per 1,000 words scanned for AIno character limits, no monthly scan caps0
Sapling5 files per batch (Free), 10 (Pro, Enterprise)From $0.005 per 1,000 charactersup to 2,000 characters per check2
Originality.aiBulk Scan1 credit/100 words ($0.01/100 words, $0.10/1,000 words)10K words (100 credits one time)3
ZeroGPTBatch files: PRO 50, PLUS 60, MAX 75, EDU 80, EXPERT 100Beginner AI Detection $0.042 / 1000 Words15,000 characters per detection, words allowance 1,250,000 /month5
ProofademicBatch Scan PDF, DOCX, TXTEssential (monthly) $15 /month, 200,000 words/moone-time 1,000 words (3-Day Trial)0
CodequiryZIP upload and bulk import of source code from Canvas, Moodle and BlackboardGrow US$200/month base (C$272 /mo as displayed): 20 web + AI checks, then C$6.80 per checkFree plan covers peer similarity only; AI detection has no free tier0
SmodinBulk processing on Premium; API billed at 0.5x the input word count, 120 requests a minutePremium $19.99/month or $16.00/month billed annually; API priced on a separate pageFree web checks without an account, 50,000 characters per check0
QuetextBulk uploads of 20 files (Essential) or 100 files (Professional); Bulk Scan up to 100 documents; URL CheckerEssential from $14.99/mo (100k words); Professional from $29.98/mo (200k words); API $0.10 per 1,000 wordsFree plan: 1,000 words per scan; 5,000 free API words0
TruthScanBatch uploads and bulk ZIP uploads in batches of 500 or fewer, CSV exportStarter $24/month, 1,000 results then $0.03 each; text API at 1 credit per word, no price per word publishedFree: 25 results a month (images and PDF pages), full API access0
IsgenBulk Scan: up to 150 files (pdf, docx, doc, 10 MB each) on Ultimate, 100 on AcademicAPI via RapidAPI, plans by contact; Academic $8 a month billed annually with 350,000 detection wordsUp to 5,000 characters per scan before sign-in; no word allowance stated0
Ircam Amplify"Maximum number of files per request: 20"; CSV batch report; 20 ISRCs per request for partners (music tracks, not text)No price per track published; pay-as-you-go from 2,000 tracks, Enterprise by contractNone published; access by application0

Values as each vendor states them, read 8 November 2026; the Codequiry row, a detector of AI-written source code, was read 9 November 2026 and shows a USD base converted to CAD by the vendor's page. Pangram, Phrasly, Sapling, Originality.ai and ZeroGPT show a dollar symbol; Phrasly, Sapling and Originality.ai name USD, and ZeroGPT's API cards are in USD while its personal plans displayed in CAD.

Bulk values come from pangram.com, the GPTZero FAQ, the Phrasly API docs, sapling.ai, originality.ai/pricing, zerogpt.com/pricing and the Proofademic detector page. Prices come from pangram.com/pricing, the GPTZero blog dated October 25, 2026, phrasly.ai/pricing, sapling.ai/pricing, originality.ai/pricing, zerogpt.com/pricing and proofademic.ai/pricing, all read 8 November 2026. The Codequiry bulk value comes from codequiry.com and its price from codequiry.com/pricing, both read 9 November 2026.

AI detectors for copywriters and agencies: register plate with the page title and its register ID
The view plate: register ID and title, drawn from the index data.

Why each card is in this view

  1. AIDI-D-0006Pangram

    API at 5 QPS realtime and uploads of up to 100 files at a time; the text API costs $0.05 per 100 words with a 20% bulk discount, and the Team plan is $40 / month for 2 seats (pangram.com, pricing, read 8 November 2026).

  2. AIDI-D-0001GPTZero

    API documented, and a free account takes texts up to 150,000 characters or a batch of up to 250 files; the Professional Team Plan carries 500,000 words per member per month with a 2-seat minimum (FAQ, pricing, read 8 November 2026).

  3. AIDI-D-0012Phrasly

    the Business API takes at least 50 words and at most 15,000 characters per request, bulk scans from a CSV of up to 100 documents, and charges $0.02 per 1,000 words scanned, inside a $100/month plan that includes $100 in API credits (API docs, pricing, read 8 November 2026). Phrasly also sells a humanizer, and its terms state "WE CANNOT GUARANTEE THAT THE TEXT GENERATED BY OUR SERVICES WILL ALWAYS PASS SUCH DETECTORS".

  4. AIDI-D-0010Sapling

    API up to 200,000 characters per request from $0.005 per 1,000 characters with a one-time $5 minimum; 5 files per batch on Free and 10 on Pro and Enterprise (sapling.ai, pricing, read 8 November 2026).

  5. AIDI-D-0004Originality.ai

    Bulk Scan and API, 1 credit per 100 words at $0.01, Pro at $12.95/month billed yearly for 200K words, added seats $8.62/month (pricing, read 8 November 2026).

  6. AIDI-D-0011ZeroGPT

    batch files from 50 on PRO to 100 on EXPERT, an API from $0.042 per 1,000 words, and a free allowance of 1,250,000 words a month at 15,000 characters per detection (pricing, read 8 November 2026). ZeroGPT also sells a humanizer.

  7. AIDI-D-0014Proofademic

    Batch Scan for PDF, DOCX and TXT, an API with 2,000 trial credits at 1 credit per word, and Essential at $15 /month for 200,000 words with a 3,500-word request limit (detector page, pricing, read 8 November 2026).

  8. AIDI-D-0029Smodin

    the Premium plan lists bulk processing and API access, and the API reference bills AI Detection V2 at half the input word count with limits of 120 requests a minute and 6,000 an hour (smodin.io/pricing, docs.smodin.io, read 9 November 2026); the web tool is free without an account up to 50,000 characters, and no study record exists.

  9. AIDI-D-0033Quetext

    bulk uploads of 20 or 100 files by plan, a Bulk Scan of up to 100 documents and a URL Checker, with an API at $0.10 per 1,000 words from a prepaid wallet (quetext.com/pricing, quetext.com/developers-api, read 9 November 2026); no study record.

  10. AIDI-D-0034TruthScan

    batch uploads, bulk ZIP uploads of 500 or fewer and CSV export on the dashboard, with a text API billed at 1 credit per word while the plans count results in images and PDF pages (truthscan.com/pricing, truthscan.com, read 9 November 2026); an agency checking text has no published price per word; no study record.

  11. AIDI-D-0020Codequiry

    API and command-line client documented, and source code submitted by paste, ZIP upload or bulk import from Canvas, Moodle and Blackboard; Grow is US$200 a month at base (C$272 as displayed to a Canadian reader) for 20 web and AI checks, then C$6.80 per check, and the free plan excludes AI detection (codequiry.com, usage/api, pricing, read 9 November 2026). Codequiry scores source code only; an agency that ships code rather than copy is the reader it fits.

  12. AIDI-D-0047Isgen

    a 150-file Bulk Scan and a RapidAPI listing with plans by contact (isgen.ai/ai-detector-bulk, isgen.ai/pricing, read 9 November 2026); 80+ languages stated and no study record.

  13. AIDI-D-0065Ircam Amplify

    the one music card in this view, with API docs and a 20-file batch ceiling for agencies that check audio rather than copy (docs.ircamamplify.io, audio input methods, read 10 November 2026); no price per track is published and access is by application; no study record. Trinka, added the same day, publishes API docs and no bulk ceiling, so it stays out; Binoculars' README batch argument is code, not a product route.

What the evidence says for copywriters and agencies

Agency copy is edited copy, and the study that tests edited text is Perkins 2024. GPTZero fell from 26.4% on non-manipulated AI text to 16.7% after adversarial techniques, and ZeroGPT from 31.3% to 17.3% with 0 false accusations in 10 human samples (Perkins 2024, arXiv, read 8 November 2026). In Weber-Wulff 2023, 13 of 36 AI documents went undetected by ZeroGPT, and 12 of 36 by Winston AI, mostly manually edited or Quillbot-paraphrased texts, while both recorded 0 false accusations in 18 human-written documents (Weber-Wulff 2023, arXiv, read 8 November 2026).

On unedited student text the picture is different. In Walters 2023, Originality.ai (model 1.4) classified 98% of 126 documents correctly with 0 incorrect, ZeroGPT 87% with a 2% false positive rate, GPTZero (Pro) 81% with 2%, and Sapling 65% with 28% returned uncertain (Walters 2023, publisher PDF, read 8 November 2026). RAID 2024 adds the attack view: Originality.ai scored 96.7% under paraphrase but 9.3% under the homoglyph attack, while GPTZero lost only 0.3% under the same attack (RAID 2024, arXiv, read 8 November 2026). For writers whose first language is not English, Liang 2023 reports an average false positive rate of 61.22% across seven detectors on 91 TOEFL essays, with GPTZero, Sapling, Originality.ai and ZeroGPT among the seven (Liang 2023, arXiv, read 8 November 2026). Pangram, Phrasly and Proofademic appear in no study record in the index.

The vendors say the same thing in their own words. Sapling states that "small modifications to AI-generated text can cause that text to no longer be flagged as AI-generated" (API docs, read 8 November 2026). No detector score should be the only evidence in a misconduct or hiring decision, and the same holds for a client dispute over a draft.

An agency that checks product reviews or other very short copy at volume has a narrower field to choose from than the one above: the marketplaces and review platforms view, added 9 November 2026, lists the two cards that state an API, a minimum under 300 characters or short-text handling, and a bulk route together, and the six that state two of the three. An agency that wants to test an API before paying has the free API question, added 9 November 2026, which lists the ten cards whose free allowance applies to API calls.

The bulk rule of this view is applied to every card in the index on the bulk and batch view, added 9 November 2026: 15 cards state a bulk upload, batch API, ZIP or folder import, the 11 here plus QuillBot and AIDetectPlus without an API and Resemble Detect and Pindrop Pulse for voice.

Sources

  • https://www.pangram.com/ (read 8 November 2026)
  • https://www.pangram.com/pricing (read 8 November 2026)
  • https://gptzero.me/faq (read 8 November 2026)
  • https://gptzero.me/pricing (read 8 November 2026)
  • https://gptzero.me/news/best-ai-detectors (read 8 November 2026)
  • https://business.phrasly.ai/docs/ai-detection-api (read 8 November 2026)
  • https://phrasly.ai/pricing?plan=monthly (read 8 November 2026)
  • https://sapling.ai/ai-content-detector (read 8 November 2026)
  • https://sapling.ai/pricing (read 8 November 2026)
  • https://sapling.ai/docs/api/detector/ (read 8 November 2026)
  • https://originality.ai/pricing (read 8 November 2026)
  • https://www.zerogpt.com/pricing (read 8 November 2026)
  • https://proofademic.ai/ai-text-detector/?utm_source=aidetectorindex.com&utm_medium=owned&utm_campaign=index_view_row (read 8 November 2026)
  • https://proofademic.ai/pricing/?utm_source=aidetectorindex.com&utm_medium=owned&utm_campaign=index_view_row (read 8 November 2026)
  • https://arxiv.org/pdf/2403.19148 (read 8 November 2026)
  • https://arxiv.org/pdf/2306.15666 (read 8 November 2026)
  • https://www.degruyterbrill.com/document/doi/10.1515/opis-2022-0158/pdf (read 8 November 2026)
  • https://arxiv.org/pdf/2405.07940 (read 8 November 2026)
  • https://arxiv.org/pdf/2304.02819v3 (read 8 November 2026)

Revision log: 2026-11: view created2026-11-09: Codequiry row added; the counts in the opening paragraph now follow the index data file at build time2026-11-09: six wave 3 cards checked against the two fields; Undetectable AI and AI or Not noted as API cards without a batch ceiling, compound count 20; 2026-11-09: ten wave 4 cards checked, Smodin, Quetext and TruthScan added, compound counts 28 and 112026-11-09: marketplaces and review platforms view linked (fifth wave)2026-11-09: free API question linked (seventh wave)2026-11-09: eighth-wave pages linked2026-11-10: Isgen added (ninth wave); the other eight ninth-wave cards state an API or a bulk route but not both2026-11-10: Ircam Amplify added (tenth wave), 13 cards; Trinka (no bulk ceiling) and Binoculars (code argument) left out

FAQ: AI detectors for copywriters and agencies

Which AI detectors fit copywriters and agencies?

Eight cards state both an API and a bulk or batch upload: Pangram, GPTZero, Phrasly, Sapling, Originality.ai, ZeroGPT and Proofademic, as read on 8 November 2026, and Codequiry, read 9 November 2026, which scores source code rather than copy. The table carries each vendor's stated batch ceiling and price so an agency can filter on volume first and on cost second.

Which detectors price per word for API checks?

Pangram at $0.05 per 100 words, Phrasly at $0.02 per 1,000 words, Originality.ai at $0.01 per 100 words and ZeroGPT from $0.042 per 1,000 words all publish a metered text rate; Sapling meters by character, from $0.005 per 1,000 characters, all read 8 November 2026. GPTZero and Proofademic price by plan with a monthly word allowance.

Which detectors take bulk uploads?

GPTZero states the largest ceiling at up to 250 files with a free account, ZeroGPT up to 100 batch files on EXPERT, Pangram up to 100 files at a time, Phrasly a CSV of up to 100 documents on the Business API, and Sapling 10 files per batch on Pro, all read 8 November 2026. Originality.ai and Proofademic name a Bulk Scan or Batch Scan feature without a stated ceiling.

Does editing AI text change the score?

Yes, in both directions. Perkins 2024 recorded GPTZero falling from 26.4% to 16.7% and ZeroGPT from 31.3% to 17.3% after adversarial edits, and RAID 2024 recorded Originality.ai dropping to 9.3% under a homoglyph attack. A draft that passed before a round of edits can fail after, and the reverse.

Which free tier covers the most words a month?

ZeroGPT states a words allowance of 1,250,000 a month on the free plan, and GPTZero states 10,000 words a month, both read 8 November 2026 (zerogpt.com/pricing; gptzero.me). The other five state per-check limits or one-time credits rather than a monthly figure, and Codequiry's free plan excludes AI detection. The free tiers ranking orders every card on this field.

Will a client's detector agree with mine?

Often not. The same text scored by two detectors on the same day can return different results, and the studies above show a single detector moving from 0 false positives in one sample to 2% in another. Record which detector, version and date produced a score before sending a draft, so a disagreement can be traced.

Related reading