Quick answer
Cursor documents max mode consumption as Token-priced; draws down included usage faster. Input, cached input, and output use model-specific request-equivalent rates.
Verified Aug 22, 2026Official source
Current limits
| Constraint | Current value | Scope | Verified source |
|---|---|---|---|
| Max mode consumptionInput, cached input, and output use model-specific request-equivalent rates. | Token-priced; draws down included usage faster | Eligible paid plans | CursorAug 22, 2026 |
Why does this limit matter?
A long-output Max-mode task can consume much more allowance than an ordinary short agent turn.
This value is scoped to Eligible paid plans; a different plan, runtime, model, endpoint, region, or account can produce a different effective constraint.
What should you check?
- Inspect the Max-mode indicator, current context size, and usage dashboard before a large task.
- 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
- Request-equivalent rates and model availability can change; use the current model table.
- 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