I’m extremely frustrated with the current limitations of the Pro plan. After hitting a rate limit over 6 hours ago, I have been completely locked out of using the Claude model entirely. I’ve attempted switching to other available models (Gemini, in particular), but these alternatives consistently fail or break my platform entirely.
Previously, hitting the 500-request threshold only resulted in slowed requests for about 5-10 minutes. This extended lockout of an entire model is unprecedented and feels excessive, particularly without any clarity or transparency about when model access might return or how long this outage will last.
The shifting rate limits and unclear pricing structure make it feel like we’re being treated as guinea pigs for your ongoing pricing experiments. It feels disingenuous, especially considering we are paying specifically for reliability and consistency.
I’d greatly appreciate clear and transparent communication regarding these issues, along with a realistictimeframe for resolution and reassurance about exactly what level of service we can expect moving forward.
Anthropic’s own Claude Code offers a quite clearer deal - you get “unlimited”, rate-limited requests (with the full context window) and access to their Sonnet models, for 20USD a month. You have about 30-40 requests for every 5 hours, that then replenish. In Claude Code, they even have a nice “You’re about to reach your rate-limit” message.
In GitHub Copilot Pro, you get 300 requests for 10$, 128k context (but they’re adding full 200k support soon-ish).
In AugmentCode you get 600 requests for 50$, with full context enabled. It uses Sonnet 4 as its “base model”, and you can’t choose any other model.
