Quick answer
Claude Code documents average max 20× prompt estimate as Approximately 200–800 prompts every 5 hours. The estimate remains workload-dependent and shares capacity with Claude.
Verified Aug 22, 2026Official source
Current limits
| Constraint | Current value | Scope | Verified source |
|---|---|---|---|
| average Max 20× prompt estimateThe estimate remains workload-dependent and shares capacity with Claude. | Approximately 200–800 prompts every 5 hours | Max 20× | AnthropicAug 22, 2026 |
Why does this limit matter?
A larger allowance supports sustained agent work but still needs model and context budgeting.
This value is scoped to Max 20×; a different plan, runtime, model, endpoint, region, or account can produce a different effective constraint.
What should you check?
- Inspect /status and the selected model before calculating how many sessions remain.
- 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
- The high end should not be treated as a promise for large codebases or long-running Opus tasks.
- 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