FAIR USE

Clear limits, no surprises.

Last updated August 27, 2026

VignetteLens is designed for normal individual study use. The safeguards below apply to the $10 API model.

What counts as a request?

On ScholarRx, Amboss, UWorld, Bootcamp, and shared Gemini quizzes, an AI request is counted only when VignetteLens calls the model. Reusing cached hints, opening a prepared tooltip, or viewing a prepared workup does not count as another request. Automatic preparation counts only on a cache miss that calls the model; a follow-up chat message counts when it calls the model.

API-model limits

• 30,000 API requests per calendar month
• $1.00 maximum estimated API cost per day
• $7.00 maximum estimated API cost per week
• $15.00 maximum estimated API cost per month
• 5,000,000 input tokens and 350,000 output tokens per day
• 12,000 input tokens and 2,000 output tokens per individual request

Conversation limits

An API question thread resets after 20 user messages, 100,000 accumulated input tokens, or 24 hours of inactivity.

What happens at a limit?

API usage is checked approximately every 30 minutes. If an API account crosses a safeguard, its dedicated credential is disabled when detected. Reactivation is manual. The VignetteLens owner must review and explicitly approve that specific account before API access is restored. Cost reporting may arrive after a request, so the safeguards reduce risk rather than guarantee exact real-time billing.

Appropriate use

Please use VignetteLens for your own educational work. Do not share account credentials, automate high-volume traffic, attempt to bypass limits, or submit real patient identifiers or protected health information. We may suspend accounts that appear to abuse the service or create security, privacy, or reliability risks.

For questions about these limits, contact vignettelens@gmail.com.