I’m a long-standing ChatGPT Pro 200 subscriber, and I am asking OpenAI for a direct, public explanation about Codex usage limits.
In the email OpenAI sent me on September 30, 2026, the company explicitly stated that existing Pro 200 subscribers would retain their current 20× Plus allowance for Codex and Work through October 29, inclusive, and that the reduction to 10× would begin on October 30.
Yet after my most recent usage reset, on October 11, my Codex allowance appears to be disappearing far faster than it did before, despite the work being broadly comparable. I am concerned that the reduced limit may already be applied to my account. I cannot verify that from the percentage meter alone, and I am not presenting it as a proven fact.
I am bringing this up publicly **because my experience of OpenAI’s private email support has been so dismissive that it feels as though they are laughing at me rather than seriously addressing my complaints.** I have already spent considerable time in a formal dispute about repeated Codex failures, wasted usage, lost value from banked resets, and unsatisfactory responses. I do not want another cycle of scripted, evasive emails.
Could someone from OpenAI please answer publicly:
1. Are existing Pro 200 subscribers actually still receiving the full 20× Codex/Work allowance until October 29, 2026, as promised?
2. Has any usage calculation, model weighting, or effective allowance already changed before that date, even if the headline multiplier is unchanged?
3. How can a subscriber verify their exact allocated quota and audit the consumption after a reset, rather than relying on a percentage?
4. If a subscriber was migrated prematurely or had usage incorrectly charged, what is the concrete remedy?
I am developing two applications. Between the quota consumption, repeated failures, duplicated work, and the way my complaints have been handled, **the current cost-to-benefit ratio makes it economically unviable for me to keep working with OpenAI tools.**
I am no longer looking for marketing assurances. I am asking for transparent figures and an actual explanation. At this rate, this is goodbye, OpenAI — for good.