# Anthropic's Claude: Faster Rate Limit Depletion During Peak Hours for All Plans
Anthropic is introducing dynamic rate limits for Claude models across Free, Pro, and Max plans. During peak hours—weekdays from 5:00 to 11:00 PT (3:00–9:00 PM MSK)—five-hour limits deplete faster. Weekly quotas remain unchanged; only the consumption rate is adjusted. For developers in Russia and Europe, this hits the second half of the workday, when AI tool demand peaks.
Anthropic engineer Tarik Shihipar, who oversees Claude Code, explained the changes as infrastructure optimization. Boosted server efficiency handles the load, but 7% of Pro users will run into the new restrictions.
Recommendations for Optimizing Workflow
To minimize impact, Anthropic recommends shifting tasks to off-peak periods:
- Resource-heavy operations (code generation, analyzing large datasets) — to evenings/nights MSK or weekends.
- Background tasks with high token usage — outside 3:00–9:00 PM MSK.
- Short queries and iterative debugging — spread evenly throughout the day.
This approach saves limits without sacrificing productivity. Meanwhile, a promo runs through April: doubled limits during off-peak hours and weekends for all plans.
Competitor Context and User Reactions
The changes coincide with OpenAI extending its promo for doubled Codex limits in ChatGPT across all plans. In discussions under Shihipar's post, developers report migrating to alternatives without such strict limits. The battle for mid- and senior-level specialists goes beyond model quality—it's about availability and quota predictability.
Peak hours target the US audience, but global teams feel the pinch. For distributed projects, this means tweaking CI/CD pipelines and AI-driven automation.
Key Takeaways
- Peak hours: 3:00–9:00 PM MSK weekdays, with accelerated five-hour limit depletion.
- 7% of Pro subscribers will hit new limits, despite infrastructure upgrades.
- Doubled limits off-peak and on weekends—through April.
- Shifting background tasks cuts peak consumption by 20–30%.
- Competitors (OpenAI) are ramping up promos, fueling user churn.
— Editorial Team
No comments yet.