Usage credits are prepaid amounts that are consumed based on your use of MindStudio's services and features. They are deducted based on the specific tasks performed, like data vectorization, API calls, or using advanced AI features. Each service deducts a different amount of credits.

If you run out of credits, any AIs that rely on usage metering will be temporarily suspended until additional credits are purchased or your monthly credits are replenished. You will receive an email notification if you're approaching your limit.

Each interaction with our AI model involves two main parts that determine the cost: the "prompt" and the "response." Prompt tokens include everything you send to the AI: your main question or command, any additional instructions, and all the data the AI uses to generate a response, like examples or results of database queries. Response tokens are what the AI sends back to you, whether it's a direct reply to your question or data it's saving for later use. A token represents roughly three-quarters of a word. So, for about every 1000 tokens, you're dealing with 750 words. Example Calculation: Imagine you send the AI a request that includes a basic prompt (500 tokens), a reference article (2000 tokens), and a content outline (500 tokens). Then, the AI produces an article that's 1000 words long. Input cost: You'll add up the tokens for the prompt, the article, and the outline. That's 500 + 2000 + 500, totaling 3000 tokens. Output cost: The article is 1000 words, roughly equivalent to 1250 tokens. Assuming you're using GPT-3.5, our default model, this interaction will cost: Input cost ≈ $0.0039 Output cost ≈ $0.0033 Total cost ≈ $0.0072