Endpoint
Response
Model routing
Themodel field in your chat completion request is mapped internally:
| Request value | Internal model | Backend | Streaming |
|---|---|---|---|
"conductor" | Mako-32B Conductor | RunPod Serverless | stream: false only |
"operator" | Mako-8B Operator | Ollama | stream: true or false |
model is omitted, the gateway defaults to the Operator model.