I’m writing as a long-time Plus subscriber to express growing frustration with the sudden and unannounced rollout of usage throttling under the $20/month GPT-4 plan.
This change—limiting usage mid-session and silently switching users to a “GPT-4 mini” fallback—has completely disrupted the workflows many of us rely on. The Plus plan was marketed as giving consistent access to GPT-4. That’s why I subscribed. That’s what I paid for.
Now I’m being capped mid-project with vague “reset in X minutes” messages and zero transparency about request limits or why the fallback even exists. This feels less like resource management and more like a quiet upsell to the $200/month Team tier, which is unreasonable for most independent creators, researchers, and working professionals.
If limits are necessary, so be it—but:
- Be transparent: Show us a usage meter or request counter.
- Be respectful: Don’t downgrade model access mid-session without warning.
- Communicate clearly: Let users know what’s changing, and why.
Right now, it feels like we’re being quietly downgraded while still paying full price.
This isn’t about getting “more for less.” It’s about getting what we paid for.
