NVHQ: Can you give several example case studies about how much the credits used?
I read your answer about how the credit usage, but still do not know how is it in real practice
Hira_PluralityNetwork
Apr 30, 2026A: Hi there, I understand why this might be causing confusion. I'll try to write a detailed answer here so that all confusions get clear.
In general, a prompt or a message takes between 10 - 100 credits.
Very simple queries -> 10 credits. Very context heavy queries spanning multiple large files -> closer to 100.
On Tier 1: You get 1667 credits per day.
This means you can do between 17-160 queries per day.
On Tier 2: You get 5000 credits per day
This means you can do between 50-500 queries per day.
On Tier 3: You get 15000 credits per day
This means you can do between 150-1500 queries per day.
If you are a power user, we recommend going for Tier 3.
Moreover, if you only use MCP and rarely use extension, your credits used will be much lower as you are using your own AI models and not relying on ours. For the MCP only use, we only deduct credits when data is inserted and we need to process your data into embeddings. Fetching data takes minimal credits.
If you still want to test first before committing, we would recommend to use our freemium version first, it gives you around 200 credits per day. You can test and see how much queries are costing you in terms of credits.