GPTZero

AI-content detector, 19M+ users, acquired by Superhuman in June 2026.

Security & Safety Freemium Has API
Researched · Published · Reviewed
RECATOOLS Score
7.8 / 10
Capability
8.5
Value for money
7.5
Ease of use
8.5
ASEAN readiness
6.5
API quality
6.5
Founded
2022
HQ
Princeton, New Jersey, USA
Users
10M+ registered users; 380,000+ educators
Launched
Jan 2023
Developer
GPTZero Inc.

Overview

GPTZero is a proprietary AI-content detector founded by Princeton's Edward Tian in 2022, with sentence-level detection across 20+ languages and LMS integrations. Superhuman (Grammarly's parent) acquired it in June 2026 for an estimated $88M, with 19M+ users and $30M ARR.

Advertisement

Pricing

Pricing shown for reference only. These figures reflect RECATOOLS research as of 11 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.

Free
$0/mo
10,000 words/month, core AI detection
  • 10,000 words/mo
  • Sentence-level highlighting
  • Browser extension
Essential
$14.99/mo
150,000 words/month for individuals
  • 150,000 words/mo
  • Faster scans
  • Basic reporting
Professional
$45.99/mo
500,000 words plus API access
  • 500,000 words/mo
  • API access
  • 250-file batches
Team
Custom
Multi-seat plans with shared credits
  • Shared team credits
  • Unified billing
  • Admin controls
Enterprise
Custom
Volume pricing for 50+ users
  • Volume pricing
  • Dedicated onboarding
  • Custom contract

Use cases

University academic integrity offices scanning student submissions for AI-generated content K-12 educators checking essays via Canvas, Moodle, or Google Classroom integration Hiring managers screening AI-written job applications and cover letters Government agencies and grant-writing offices verifying authentic human authorship AI training data labeling teams certifying that datasets contain genuine human text

What you can produce with GPTZero

  • Per-sentence AI probability highlighting across a scanned document
  • Overall AI vs. human confidence score with percentage breakdown
  • Plagiarism detection report (Premium plan and above)
  • Writing feedback and grammar analysis overlay
  • Document replay showing the writing history of a student submission
  • LMS-native submission scanning via Canvas, Moodle, Blackboard integrations
  • API returning per-sentence and document-level AI scores (paid tiers with word-allotment limits; custom/enterprise for highest throughput)
Advertisement

ASEAN Perspective

GPTZero in Southeast Asia

GPTZero supports Chinese (Simplified/Traditional), Korean, and Japanese with claimed near-zero false positive rates, which is relevant for ASEAN universities using CJK-language instruction. However, no ASEAN-specific institutional partnerships, regional pricing, or Southeast Asian-language (Malay, Indonesian, Thai, Vietnamese) detection have been publicly announced as of mid-2026. For Singaporean and Malaysian institutions, the platform's FERPA/GDPR compliance posture translates reasonably to PDPA obligations, but procurement teams should conduct independent false-positive benchmarking on local student writing before institutional rollout.

RECATOOLS Verdict

GPTZero's June 2026 acquisition by Superhuman -- the productivity company Grammarly formed after buying the email client of the same name -- confirms what its metrics already suggested: 19M+ registered users, $30M ARR, and a deal PitchBook values above $88M against just $13.5M raised. Both founders and about 30 staff joined Superhuman; the product keeps running standalone while also feeding Superhuman's Go AI assistant, so near-term continuity looks solid. Sentence-level highlighting, 20+ language coverage, and Canvas/Moodle/Blackboard integrations still lead the education category, and 2025's Model 3.7b update covers GPT-4.1, o3, Gemini 2.5 Pro, and Claude Sonnet 4 without retraining.

The accuracy gap hasn't closed: independent testing keeps landing at 6-20% real-world false positives against GPTZero's own 99%+ lab claims, and several US universities have pulled AI detectors from disciplinary use over wrongful accusations. API access still sits behind paid tiers.

Independent AI-assisted assessment by RECATOOLS.

What people say

Nineteen million registered users and $30 million in ARR were enough to get GPTZero bought. Superhuman -- the productivity company formed when Grammarly acquired the email client Superhuman and took its name -- announced the deal on June 23, 2026, with both GPTZero founders and roughly 30 staff joining the buyer. PitchBook pegs the transaction above $88 million against just $13.5 million raised across a 2023 seed and a 2024 Series A, a genuinely capital-efficient exit for the AI-tools space.

For now the product keeps running as a standalone brand rather than getting folded in outright, with plans to also surface inside Superhuman's Go AI assistant. That continuity matters to the roughly 380,000 educators and institutions running Canvas, Moodle, and Blackboard integrations, plus the American Federation of Teachers partnership covering 1.7 million members. Business Insider named it the best free AI detector in 2025, and a 2025 model update extended coverage to GPT-4.1, o3, Gemini 2.5 Pro, and Claude Sonnet 4 without explicit retraining on those models.

The accuracy gap hasn't closed, though. GPTZero claims 99%+ accuracy under lab conditions; independent field testing keeps landing in the 6-20% false-positive range, and that spread has already pushed some US universities to pull AI detectors from disciplinary use entirely. API access still sits behind paid Professional and Team tiers starting around $46/month, so solo developers are locked out. No confirmed Southeast Asian-language benchmarking or ASEAN partnerships exist yet, and it's an open question whether Superhuman keeps investing in GPTZero as its own product or folds the detection tech into email tooling over time.

Summary of public user & expert reviews, compiled by RECATOOLS.

Notable facts

  • GPTZero crashed its Streamlit-hosted server on launch day (January 2023) after 30,000 users signed up within the first week.
  • Co-founder Edward Tian built the original prototype as his Princeton senior thesis project over a winter break.
  • GPTZero's summer 2025 model detects Claude Sonnet 4 outputs at 99.1% recall — despite not being explicitly trained on Claude data.
  • The American Federation of Teachers deal gave GPTZero access to 1.7 million educators in a single institutional contract.

Frequently asked questions

Is GPTZero accurate enough to use for academic misconduct proceedings?
GPTZero claims 99%+ accuracy under controlled conditions, but independent studies and real-world cases document false positive rates of 6–20%, with higher rates for non-native English writers. Multiple universities have restricted AI-detection tools after wrongful-accusation incidents, including lawsuits at Yale (2025). GPTZero should be treated as one signal in an investigation, not definitive proof of misconduct.
Does GPTZero work in languages other than English?
Yes. As of October 2025, GPTZero's multilingual model covers 9+ languages including Chinese, Korean, Japanese, Arabic, Spanish, French, German, Portuguese, and Italian, with benchmarks across 20+ languages. GPTZero's own data shows near-zero false positive rates for Chinese and Korean, though independent third-party validation for Asian languages is limited.
Is the API available on the free or individual paid plans?
No. API access is restricted to Institutional and Enterprise plans, which are custom-priced. Individual plans (Essential at ~$8.33/month to Professional at ~$45.99/month) do not include API access. Developers looking to integrate AI detection must contact GPTZero's sales team for institutional/enterprise pricing.

About this listing

Researched on
Published on
Last reviewed

This entry was compiled from publicly available data including GPTZero's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with GPTZero unless explicitly stated.

Data accuracy

Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.

For the latest details, please refer to GPTZero directly →

Spotted something out of date? Suggest an update →

Advertisement