—
Total Requests
—
last 7 days
—
Total Cost
—
USD
—
Total Tokens
—
prompt + completion
—
Avg Latency
—
end-to-end response time
—
Error Rate
—
of all requests
—
Cache Hit Rate
—
requests from cache
System Health
Database
Checking…
Routing Strategy
—
Circuit Breakers
All closed
Active Keys
—
Requests Over Time
Request volume across all providers
Cost by Provider
Cost Over Time
USD spend per day
Tokens by Model
Hourly Request Volume
Requests per hour (last 7 days)
Token Usage Over Time
Latency Percentiles by Model
P50, P90, P99 in milliseconds
Period Total
—
Cost / Request
—
Cost / 1K Tokens
—
Projected 30d
—
Daily Cost Breakdown
Model Cost Breakdown
| Provider | Model | Requests | Tokens | Cost (USD) | Cost / 1K tok | Avg Latency |
|---|
User Cost & Usage
| User | Requests | Tokens | Cost (USD) | Budget | Utilization | Avg Latency |
|---|
Budget Status (Current Month)
| User | Budget (USD) | Spent (USD) | Remaining (USD) | Utilization |
|---|
Model Performance & Cost
| Provider | Model | Requests | Prompt Tok. | Compl. Tok. | Cost (USD) | P50 | P90 | P99 |
|---|
Pricing Reference
| Provider | Model | Input ($/1M tok) | Output ($/1M tok) |
|---|
Per-User Quota Status
Real-time counters — reset at period boundary
| User | Period | Resets | Tokens Used | Token Quota | Token Util% | Requests Used | Request Quota | Req Util% | Status |
|---|
Quota Configuration
Set per-user quotas via the Admin API. Setting either limit to
0 disables that dimension.
quota_period accepts "daily", "weekly", or "monthly".
curl -X PATCH http://localhost:8000/admin/users/<user_id> \
-H "x-admin-key: $ADMIN_KEY" \
-H "Content-Type: application/json" \
-d '{
"token_quota": 1000000,
"request_quota": 500,
"quota_period": "monthly",
"quota_enabled": true
}'
| Time | Provider | Model | Tokens | Cost | Latency | Status | Cache | Stream |
|---|
Virtual API Keys
| Name | Key Prefix | Provider | Routing | RPM | TPM | Daily Limit | Last Used | Status |
|---|
Provider Keys
| Provider | Name | Organization | Weight | Priority | Health | Last Used | Status | Actions |
|---|
Add Provider Key
Pool Policy
Only applies once an organization/provider has 2+ active keys — controls how requests are spread across the pool and what happens if every key is cooling down after failures.
Community
Usage vs License Limits
Licensed Features
Manage License
License activation and removal are managed from the
Admin Console.
Tier Comparison
| Feature / Limit | Community | Professional | Enterprise |
|---|
Distributed Traces
| Trace ID | Spans | Model | Provider | Cost | Duration | Status | Started |
|---|
Total Requests
—
Cache Hits
—
Hit Rate
—
Cost Savings
—
Hit Rate by Model
Cache Breakdown by Model
| Model | Provider | Requests | Cache Hits | Hit Rate | Savings (USD) |
|---|
Cache Breakdown by User
| User | Requests | Cache Hits | Hit Rate | Savings (USD) |
|---|
MCP Gateway Traffic
Read-only usage across upstream tool calls and the inbound gateway. To register or configure MCP servers, use the Admin Console.
Calls
—
Cost
—
Input Tokens
—
Output Tokens
—
Breakdown by Tool
| Tool | Calls | Cost | Success | Error Rate | P50 | P95 |
|---|