Appearance
Manage Token Settings
Token settings allow you to control how your API keys consume resources. You can set quotas, rate limits, and assign keys to token groups for organized management.
Token Quota
Each API key can have an optional token quota — the maximum number of tokens it can consume. Once the quota is reached, requests using that key will be rejected until the quota is reset or increased.
Setting a Quota
- Go to your dashboard
- Navigate to Token Management
- Select the key you want to configure
- Set the Token Quota value
- Save changes
Quota Reset
Quotas can be configured to reset on a schedule (daily, weekly, monthly) or managed manually from the dashboard.
Rate Limits
Rate limits control the maximum number of requests a key can make within a time window.
| Setting | Description |
|---|---|
| Requests per minute | Maximum API calls per minute |
| Tokens per minute | Maximum tokens consumed per minute |
| Requests per day | Maximum API calls per day |
Token Groups
Token groups let you organize multiple keys under a single management unit. This is useful for:
- Team management — Group keys by team or department
- Project isolation — Separate token budgets per project
- Usage tracking — Monitor consumption at the group level
Creating a Token Group
- Go to Token Management → Groups
- Click Create Group
- Set the group name and shared quota
- Assign API keys to the group
Modifying Token Settings
You can modify token settings at any time from the dashboard:
- Increase/decrease quotas — Adjust based on actual usage patterns
- Change rate limits — Tighten or relax as needed
- Move keys between groups — Reorganize without recreating keys
- View usage history — Check consumption trends over time
Usage Monitoring
The dashboard provides real-time usage statistics:
- Total tokens used — Across all keys or per key
- Request count — Number of API calls
- Model breakdown — Token usage per model
- Time series — Usage trends over time
TIP
Set up quota alerts to get notified before you hit your limits. This prevents unexpected service interruptions.