> faq --overview
> faq.billing_model()
How do subscriptions map to pages and credits?
Paid plans run on a monthly credit budget. The pricing cards show lecture-equivalent pages because lecture processing is the heaviest pipeline and uses the full 5x multiplier.
├─ upload tier: 1 credit per page
├─ summarize tier: 4 credits per page
└─ lecture tier: 5 credits per page
That means lighter processing tiers cover more total pages in the same billing period. For example, a micro plan can cover about 200 lecture pages, 250 summarize pages, or 1,000 upload-only pages in a month.
What is included in each processing tier?
├─ upload: PDF parsing and storage only
├─ summarize: summary plus quiz generation
└─ lecture: summary, quiz, lecture, and audio generation
How does the free tier work?
Public browsing stays free. New accounts also get 10 welcome unlocks for public papers. Those welcome unlocks are separate from paid subscription credits.
Do I need to verify my email before paying?
Yes. You can still use the free parts of the product without verification, but paid checkout stays locked until your email is confirmed.
> faq.billing_ops()
Who sends receipts and subscription billing emails?
Paddle sends receipts, invoices, and subscription lifecycle emails for paid plans. Dissecting AI currently sends account emails such as verification and password reset separately.
How do cancellations work?
You can manage or cancel a subscription through the billing portal linked from the app. In normal cases, cancellation stops renewal and your current paid period remains available until it ends.
Do you offer refunds?
Refunds are reviewed manually. Partial refunds may be granted case-by-case based on unused credits or unused billing value. There is no automatic in-app proration or guaranteed refund window.
> faq.data_location()
Where is data processed?
Dissecting AI is designed so application-controlled processing should remain in the European Union.
Third-party providers were selected in part because they represent that relevant processing happens in the EU. As a practical matter, however, we rely on those provider statements and cannot independently guarantee every aspect of their infrastructure or operations at all times.
Which providers are involved?
Depending on the feature, the product uses Paddle for billing, Brevo for transactional email, PostHog for analytics, Mistral and Nebius for AI processing, and Bunny.net for storage or delivery. The privacy page describes these categories in more detail.
Where should I send billing, refund, or privacy questions?
Email [email protected].
> faq.content_quality()
How reliable are summaries, quizzes, lectures, and other paper outputs?
Most paper-related outputs are AI-generated and may not be 100% accurate. They should be treated as high-quality study aids, not as a substitute for checking the source paper directly.
In day-to-day use, the product has held up well for the author, who uses it heavily, but occasional mistakes, omissions, or awkward phrasing can still happen.
Does the same caveat apply to TTS audio?
Yes. The same applies to lecture audio and other TTS output. In practice it has been working well, but minor glitches such as odd pronunciations, timing issues, or brief audio artifacts may still occur.
What should I do if I spot a content or TTS issue?
If you notice a clear content or audio issue, contact [email protected] so it can be reviewed.