Free
For occasional checks.
Free
Sign up free - 10 detections per month
- Up to 1,000 words per detection
- Free signup, no credit card
- Standard AI detection model
- 7-day detection history
Simple, transparent pricing
Industry-leading AI detection trained on 2 billion+ samples. Free tier real-detection. Paid tiers from $9. Or pay only when you need it with one-time word packs that never expire.
Starter
For solo creators.
$9 /mo
Get Starter - 1,000 detections per month
- Up to 2,000 words per detection
- Per-sentence AI/human verdicts
- Full advanced detection model — trained on 2B+ samples
- Forever detection history
Most popular
Pro
For regular users.
$19 /mo
Get Pro - 2,500 detections per month
- Up to 2,000 words per detection
- Everything in Starter
- PDF + JSON export reports
- Strictness toggle (lenient / balanced / strict)
Plus
For longer-form content.
$39 /mo
Get Plus - 3,500 detections per month
- Up to 4,000 words per detection
- Everything in Pro
- REST API access (coming soon)
- Counterfactual analysis — how to rewrite to score human
Business
For teams.
$99 /mo
Get Business - 5,000 detections per month
- Up to 5,000 words per detection
- Everything in Plus
- Bulk upload + batch processing
- 5 team seats
- Priority processing — no waiting
- REST API access (coming soon)
- Email support
Enterprise
For agencies + publishers.
$499 /mo
Get Enterprise - 10,000 detections per month
- Up to 20,000 words per detection
- Everything in Business
- 25 team seats
- Webhooks + custom integrations (coming soon)
- REST API access (coming soon)
- SLA + dedicated support
- Custom contract available
Pay-as-you-go
No subscription. Top up your balance ($10, $50, $100, $500, $1,000) and consume credits at $0.10 per detection (up to 5,000 words). Best for occasional or sporadic use. Start PAYG →
FAQ
What does "detection" mean?
One text submission within your tier's word limit, scored by our advanced AI detection model. Per-sentence verdicts on Starter and above. Real-time analysis — sub-second response on most documents.
What if I exceed my monthly limit?
You can either upgrade or top up via PAYG ($0.10/detection). No automatic overage charges. The dashboard shows your usage in real time so there are no surprises.
Can I cancel anytime?
Yes. Cancel from your dashboard — access continues until end of current billing period. No questions asked, no retention call.
Do you offer refunds?
14-day money-back guarantee on any paid subscription. Email [email protected] with your account email and we'll process it within one business day.
Why is your pricing lower than GPTZero or Originality?
We pass through the savings instead of marking up the margin. Industry-leading accuracy at every volume tier without the legacy-detector premium. The pricing grid above is the whole offer — no upsells, no per-token surprises.
Is there an API?
REST API access is coming soon on Plus, Business, and Enterprise tiers. The web dashboard, exports, and per-passage highlights all work today. Preview the API docs → or drop your email to be notified when API access goes live.
What are word packs?
One-time, never-expire detection credits. Buy a pack ($5 - $500), credits land in your account immediately, and they get consumed when you run detections. They stack on top of any subscription tier — when your monthly quota is exhausted, credits kick in automatically — and they work standalone on the free tier too. The bigger the pack, the lower the effective per-word cost.
Do word-pack credits ever expire?
Never. Once purchased they stay in your account until consumed. If you let the subscription lapse the credits remain available — sign back in any time and they're still there.
What models can Deep AI Detector catch?
Trained on 2 billion+ samples covering ChatGPT, Claude, Gemini, Llama, Mixtral, DeepSeek, Qwen, GLM, and every other major LLM family released in the last three years. Plus paraphrased and humanized AI text that legacy detectors typically miss.
How is my data handled?
Submitted text is deleted within 24 hours unless you explicitly opt into detection history. We never train models on your text. Hash-based caching means even our cache never stores plaintext. See the privacy policy for the full breakdown.