Quota & Rate Control

Understand the three-tier quota system and configure Key-level rate limits and Token usage for precise API control

This page covers Key-level quota and rate configuration. Each Key can have independent RPM, TPM, and Token limits. Configure company-level defaults in System Settings → Quota Management.

Three-Tier Quota System

Tarogo AI Gateway quota management is organized into three tiers, from company to individual:

🏢 Company Quota (Tenant Level)

The company quota is the highest-level default quota, applied to all users and departments. When no custom quota is configured for a user or department, the company quota is used automatically. Configure in System Settings → Quota Management.

Quota Management →

🏛️ Department Quota

Department quotas allow setting independent limits for specific departments. All users in the department inherit the department quota by default. Configure in Department Management; leave empty to inherit the company quota.

Department Management →

👤 Personal Quota

Personal quotas are individually configured for specific users with the highest priority. Configure in User Management; leave empty to inherit the higher-level quota.

User Management →

Quota Priority & Fallback

Personal Quota > Department Quota > Company Quota: the effective quota follows this priority
When personal quota is empty, it falls back to the department quota
When department quota is empty, it falls back to the company quota

💡 Tip: When editing quotas in User Management or Department Management, leave the field empty to clear the current tier quota — the system will automatically fall back to the higher-tier quota.

Rate Control

Rate control limits API Key request frequency to prevent burst traffic from overwhelming backend services. The system enforces rate limits through two dimensions: RPM and TPM.Configure rate params when creating Key →

RPM

RPM (Requests Per Minute): Max API requests allowed per minute

TPM

TPM (Tokens Per Minute): Max Token consumption allowed per minute

RPM and TPM form a dual rate limit — exceeding either threshold triggers rate limiting (HTTP 429). Together, they control traffic from both request count and request size dimensions.

Rate Limiting & Alerts

Rate Limit Response

Returns HTTP 429 when RPM/TPM exceeds limit.

Quota Alert

Status turns orange at 80% usage.

Usage Statistics

Shows usage/daily quota with percentage bar.

Reset Cycle

Daily: 00:00. Monthly: 1st at 00:00.

Important Notes

Company quota changes only apply to new Keys; existing Keys need individual editing
Monthly quota > daily quota
Set RPM/TPM based on business needs
Quotas are in Token count
Personal and App Key quotas are per-user
Check usage in real time from Key list