Credits Explained
For free tier users who have a monthly quota limit, it is important to understand that one API request does not always equal one credit. The cost of a request depends on the complexity of the endpoint you are hitting.
Base Cost: V1 Endpoint
2.20.1The standard V1 endpoint (/api/v1/q) is the baseline. Regardless of whether you fetch 1 question or the maximum limit of 50 questions in a single batch, calling the V1 endpoint consumes exactly 1 credit. This makes batching incredibly efficient for your quota.
Multiplier Costs: V2 and V3
2.20.2Because the V2 (Metadata) and V3 (Explanations) endpoints require deeper database queries and statistical aggregation, they carry a multiplier cost. A single request to V3 might cost 2 or 3 credits per call, depending on current server load and your specific free-tier rules.
AI Inference Costs: V4
2.20.3The V4 endpoint generates dynamic hints using a Large Language Model. Because AI inference requires expensive GPU compute time, hitting the V4 endpoint consumes significantly more credits per request. Building a heavily trafficked app on V4 will almost certainly require upgrading to an unlimited paid plan.
Free Endpoints
2.20.4Utility endpoints, such as fetching the list of subjects (/subjects) or reporting a typo (/report), are completely free. They do not consume any of your monthly quota credits, so you can call them as often as your UI requires.