I’ve been a long-time ChatGPT Plus subscriber and heavily relied on the DALL·E image generation feature.
In recent weeks, image quality has significantly dropped — texture, lighting, background detail, and composition are all worse than before.
OpenAI keeps responding with: “Use quality: ‘hd’”.
But as a ChatGPT web user, I cannot manually set this.
And even when using advanced prompt strategies, the results are clearly inferior to previous outputs — just slower, not better.
No transparency, no real answers.
Were all users silently downgraded? Is this an A/B test? Why are we not being told the truth?
If you’re experiencing the same, please speak up.
We’re paying for a service that is no longer delivering what it promised.
