Custom Backend Overview | NVIDIA Dynamo Documentation

Dynamo supports custom backends through one preferred unified contract, a lower-level worker path, and a packaging path:

Path	Use when
Writing Unified Backends	You are writing a new token-in-token-out engine in Python or Rust and want Dynamo to own the runtime lifecycle.
Python Workers (lower-level)	You need direct control of `register_model`, `serve_endpoint`, or the request payload.
Runtime Containers	You need to package a built-in or custom backend into a deployable Dynamo image.

The unified backend path is the preferred starting point for new custom engines. It gives Python and Rust backends the same lifecycle shape: parse arguments, start the engine, stream generated chunks, handle cancellation, drain, and clean up. The Dynamo framework owns runtime registration, signal handling, model registration, and graceful shutdown.

Use the lower-level Python worker path when your backend must own model registration, endpoint serving, request handling, or lifecycle behavior.

If your custom engine uses KV-aware routing, publish KV events so the Dynamo router can track which workers hold each prefix.