I am encountering confusion regarding my free tier token usage and the billing on my account. Below is a summary of my situation:
Problem:
I have opted in enable for all projects , and I do see the following message in “Free usage of up to 11 million tokens per day” , as stated in the UI: “Get free usage of up to 1 million tokens per day across gpt-4.5-preview, gpt-4o and o1, and up to 10 million tokens per day across gpt-4o-mini, o1-mini, and o3-mini on traffic shared with OpenAI through April 30, 2025.”
Despite using the listed models under the free tier, I received a bill of $58. Additionally, my requests are being rejected with the error:
statusCode: 429
error: "You exceeded your current quota, please check your plan and billing details."
This error message suggests there may be an issue with my free token quota being incorrectly tracked or applied.
The UI should clearly indicate whether my usage is being counted as free or if I’m exceeding the free tier limits.
There should be a clear differentiation between usage of free tokens and billed tokens in the dashboard.
A revised UI update could display a specific breakdown of how many free tokens have been consumed versus billed tokens to avoid confusion.
Additionally, it would be helpful if the UI could display a warning when a user is close to exceeding their free token limit, so they can avoid errors like statusCode 429.
Question:
Given that I’ve opted in for the free tier with the models listed in my quota, and I see the message “You’re enrolled for up to 11 million complimentary tokens per day”, should I be paying the $58 I see on my bill, or is this a mistake due to incorrect token usage tracking? How can I fix this issue and ensure my free tokens are properly applied?
The main usage page shows you what you were billed.
Complementary tokens simply don’t show up there. They aren’t billed (when everything is working right).
Checking Your Full Token Consumption
To see your full token consumption, you will need to go to the legacy dashboard, pick activity, and then select all the qualifying models as a group.
Then you get a better look at a day’s consumption—input and output tokens by day, stacked by model. This view is unfiltered. There is no “free” notation there, either.
Understanding Promotional Token Limits
The promotional consumption you are allotted is windowed, though. It doesn’t reset at a specific time. More free use is possible when 24 hours have passed since individual usages (and maybe not exactly tied to individual calls).
So you can create your own 24-hour period that never exceeds 1 million tokens on o1 models. Or—you can mistakenly start another 1 million tokens of usage at some arbitrary midnight within less than 24 hours, and then be billed for all of them.
Errors When Out of Credits
The error reported is simply that you are out of credits. Whether the tokens were free or not, your account is shut off until the API credit balance is made positive again.
OpenAI’s Intentions Regarding Free Tokens
I don’t know that OpenAI’s goal would necessarily be to maximize the amount you can submit for free and then stop (which is what a clear display of “remaining free tokens” might encourage). Instead, OpenAI is paying for your training data to understand normal and useful API usage patterns.
Any information about free tokens would also be hard to understand clearly without fully comprehending the token system described above. It would be even harder to misread than a credit balance showing $0.00.
Investigating Possible Billing Errors
Your usage bar seems to have maxed out in one day. You can reconcile daily tokens versus daily billings to see if the system accidentally broke on your account. Generally, the system hasn’t broken—I have three days of usage that were billed only for DALL-E.
I also had this same issue with both GPT-4o and GPT-4.5, I had unexpected billing even though I was well under (~500k tokens) the 1m tokens for GPT-4o for the day.
