You're right that different AI models consume words (or tokens) differently, and that's exactly why we show the word consumption rate next to each model — so users can choose based on their needs and available credits. Models like Claude 4 or GPT‑4.5 are incredibly powerful but also significantly more expensive in token usage.

We’re working on ways to better normalize consumption across models and improve transparency. Our goal is to give users both flexibility and clarity, whether you're using our credits or your own API.

Thanks again, and we hope you'll give Followr another try — we’re constantly improving based on feedback like yours