A: Hi Matteo,

Sure, I’d be happy to walk you through how you can make the most of Diaflow’s features. You can absolutely use your Diaflow credits to run LLM models, or if you prefer, leverage your own API key to integrate LLM models through our Bring Your Own Key (BYOK) option—giving you plenty of flexibility to tailor it to your needs.

Here’s how it works in Diaflow: When you want to scrape a webpage, you’ll use our Web Scraper node, which costs 20 credits per scrape. This doesn’t tap into OpenAI credits directly—it’s all handled through your Diaflow credit balance. Now, if you decide to process that scraped data using OpenAI’s models (hosted on our cloud - ChatGPT 4o), the pricing breaks down like this: it’s 0.055 credits per input token and 0.022 credits per output token. Let me give you a quick example to make it clear—say you scrape a page (20 credits), then analyze 100 input tokens (that’s 5.5 credits) and generate 100 output tokens (2.2 credits). That would come to a total of about 27.7 credits. Keep in mind, though, the exact credit usage for the OpenAI node will depend on the length of your prompt—so the more concise or detailed you go, the more it might vary.

I hope that helps clarify things!

Best,
Diaflow team