Quick answer
Claude documents usage reset window as Rolling five-hour window. Claude and Claude Code subscription activity can draw from the same window.
Verified Aug 22, 2026Official source
Current limits
| Constraint | Current value | Scope | Verified source |
|---|---|---|---|
| usage reset windowClaude and Claude Code subscription activity can draw from the same window. | Rolling five-hour window | Pro and Max | AnthropicAug 22, 2026 |
Why does this limit matter?
A burst of complex work can exhaust capacity early even when a calendar-day count looks low.
This value is scoped to Pro and Max; a different plan, runtime, model, endpoint, region, or account can produce a different effective constraint.
What should you check?
- Read the in-product remaining-capacity warning and its displayed reset time.
- Confirm the exact plan, model, runtime, endpoint, region, and account that serve the failing workload.
- Record the observed value, response headers or configuration, timestamp, and source without logging secrets.
Important caveats
- Anthropic presents prompt counts as estimates because task length and model compute alter consumption.
- Treat the official source and live account configuration as authoritative if they differ from this verified snapshot.
HyperObserve reports the documented platform constraint. Your application, SDK, gateway, provider, region, or account can impose a lower effective limit.
Related