Q: How do the credits work?
David_SheetXAI
Dec 16, 2025A: Hey c4d80a5acc7a4a27a278d4c4ab146aec 👋!
So here's how the credits work:
Each tier comes with AI credits that let you use our API keys instead of bringing your own. These are one-time credits (no monthly renewals - we keep it this way to make the LTD sustainable).
When you use SheetXAI with our credits, you're accessing most AI models through our API keys. Different operations use different amounts of credits depending on the AI model you're using and what you're doing (text generation, image generation, video, audio, etc.).
Once you run out of credits, you have two options:
1. Purchase additional AI credits as an add-on
2. Switch to BYOK (Bring Your Own Key) and use your own API keys with providers like OpenAI, Claude, Gemini, Straico, OpenRouter, Fal AI, and many more
You can also start with BYOK from day one if you prefer and save your credits for later, or mix and match however you want. It's totally flexible.
The credits never expire.
Does that answer your question or do you want more details on specific credit usage?
Let me know if you have other questions! 🙂
Thanks, but what does one credit provide? Is it like 1 credit per operation?
Verified purchaser
Hey c4d80a5acc7a4a27a278d4c4ab146aec 👋!
Unfortunately it varies a lot depending on the model you choose. Here is what I can tell you:
The formula is: Base cost X Model cost
Chatbot operations have a base of 5 credits. So if you choose a model that charges 2 credits (like a cheaper model such as GPT-4o mini), then each chatbot operation would charge you 10 credits (5 X 2).
Verified purchaser
For the cell by cell operations (SAI commands), the base is 1. So if you choose a model that charges 2 credits it would do 1 X 2 which is 2 credits per row.
The credit cost per model varies based on the AI provider and model you're using. More powerful models like Claude cost more credits per operation than lighter models like GPT-4o mini.
Verified purchaser
You can see the credit cost in the settings when you select your models.
Does that help clarify it?
Let me know if you have other questions! 🙂