Total requests
—
Connect to load statistics
Live gateway overview
One control plane for every model.
Route, observe, and test your private multi-provider LLM stack without leaving the gateway.
Success rate
—
No request data
Tokens processed
—
Input + output
Average latency
—
Current inflight load
Provider health
Priority, load, circuit state, and latency
| Provider | Status | Protocol | Models | Latency |
|---|---|---|---|---|
| Connect with an admin token to load providers. | ||||
Traffic distribution
Successful requests by provider
Provider usage appears after requests complete.
▶
Ready to test a route?
The playground automatically loads providers, discovered models, and every common generation control.
Request laboratory
Playground
Compose production-shaped requests with live provider and model configuration.
Configuration
Waiting for connectionProvider configuration loads after connect
Advanced generation settings temperature · top_p · max_tokens · stream
Stream response
Response
Connect, choose a model, and run your first request.
Request preview
{}Routing inventory
Providers
Live configuration and health for every enabled or standby upstream.
Provider registry
Connect to inspect provider configuration.
| Provider | Status | Base URL | Priority / weight | Models | Requests | Latency |
|---|---|---|---|---|---|---|
| No provider data loaded. | ||||||
Operational analytics
Statistics
Request volume, reliability, token throughput, and provider traffic from the running process.
Successes
—
Completed requests
Failures
—
Failed attempts
Failovers
—
Automatic reroutes
Configured models
—
Discovered exact model IDs
Provider request share
Distribution across successful upstream calls
No traffic statistics loaded.
Runtime snapshot
Current process-level indicators
Connect to load runtime statistics.