InferenceApiServer (Javalin) is the single HTTP entry point on the coordinator. It exposes the
Juno-native surface below alongside the OpenAI-compatible API. Both
surfaces share the same underlying RequestScheduler and GenerationLoop, so behavior and
performance are identical regardless of which surface a client uses.
Native inference endpoints¶
| Method | Path | Handler |
|---|---|---|
POST | /v1/inference | handleBlockingInference: blocking, returns GenerationResult |
POST | /v1/inference/stream | handleStreamingInference: SSE, one event per token |
GET | /v1/models | OpenAiChatHandler.handleListModels |
GET | /v1/models/{modelId} | OpenAiChatHandler.handleGetModel |
DELETE | /v1/models/{modelId} | handleUnloadModel |
GET | /v1/cluster/health | handleClusterHealth: per-node health rollup |
Health and console¶
| Method | Path | Handler |
|---|---|---|
GET | / | handleConsole: embedded coordinator web console |
GET | /health-ui | handleHealthDashboard: node health dashboard HTML |
POST | /health/probe | handleHealthProbeProxy: proxies probe to HealthReporter |
GET | /health-data | handleHealthDataProxy: proxies health JSON from nodes |
See also¶
<- 4.8 Testing Checklist | Table of Contents | 5.2 OpenAI-Compatible API ->