Using Credits vs Using Own API keys - (video lesson) 🎓

A lot of users of AI products ask whether they should use credits ortheir own API keys, and even more commonly what either of those two things means! Here’s a detailed video discussing using AI credits vs API keys. The ideas in video are applicable not just to Pickaxe, but probably a wide array of AI products and AI platforms out there, as credit/API-key paradigms are very common pricing models for AI companies.

Most AI platforms are paying for usage of AI models via API. And then pass along the cost to users via the abstraction of “a credit”. In reality, these AI platforms are paying per-token, so the costs are not fixed. They’re highly variable. It’s like pumping gasoline at a station. But that’s a complicated concept to pass along to users and a speed bump against a simple user experience. So platforms often simplify it into uniform units (i.e. credits). I hope this video is informative to anyone wondering about the question.

3 Likes

If I brought my own OpenAI key, but my Studio meters chatbot use by Credits, how is that managed by Pickaxe? Is the user still metered via Credits? How is a Credit calculated in that case? Thank you

Thanks for asking! If you use your own OpenAI API key, you can set your bot to Unlimited credits. In that case, Pickaxe credits aren’t used at all, and your OpenAI usage is billed directly to your API key. If you do choose to meter, it’s simple: 1 credit = 1 interaction (one AI reply).

Thank you for explaining!