AI credit usage seems very high
I'm not sure what's happened in the last few weeks, but my credits seem to be SHOOTING down.
I just topped up 10 credits, asked Sonnet 5 Medium “tell me what my uncommitted files are doing” (I haven't opened the project in a long time) and it thought for 1s, worked for 5s, thought for 1s and my credits had dropped to 9.80 credits.
At this rate, I'm going to get only 4 or 5 full conversations for £10.
Is there any way we can monitor the usage?
This feels like a lot of credits for very little AI.
Please sign in to leave a comment.
Hello Dave,
Thank you for your feedback, and I apologize for the experience you had.
Credit usage is calculated dynamically for each interaction based on several factors, primarily the number of tokens used — units of text from both your input and the AI’s response. Since credit usage can vary depending on the length, complexity, and context of a request, the credit cost may differ from one interaction to another.
However, in some cases, there could be other reasons for high quota consumption. To understand and analyze the root cause in your case, I request that you open a ticket with our support team here: https://www.jetbrains.com/support/, and share some details so that we can assist you with the issue.
You can alos check the following to understand more about credit usage: https://helpserver.labs.jb.gg/help/ai-assistant/2025.2/licensing-and-subscriptions.html
https://youtrack.jetbrains.com/articles/SUPPORT-A-2448
Warm regards,
Sajal
Seems like the credit system is costing more per query than expected. Likely because each request is deducted based on processing time and complexity, not just the length of the answer. Most platforms have a usage dashboard or billing history that allows you to closely monitor your consumption, so check for monitoring tools in your account settings or billing page. If that’s not feasible, the best course of action is to reach out to support directly to confirm how credits are being calculated and if your pattern of usage can be optimized.
Best regards,
Albert
AARP Membership
Thanks for the explanation. I understand that token-based pricing means the cost can vary, but I think what many users are missing is visibility into why a particular prompt consumed so many credits. It would be really helpful if the IDE showed a per-request breakdown (or at least an estimate) of credits/tokens used. That way users could tell whether a large project context, attached files, or a long conversation history is driving the cost, instead of feeling like credits are disappearing unexpectedly. I think that kind of transparency would go a long way toward addressing concerns like Dave's.
I checked the link to understand more about credit usage: https://helpserver.labs.jb.gg/help/ai-assistant/heardle/2025.2/licensing-and-subscriptions.html