> For the complete documentation index, see [llms.txt](https://docs.seldon.ai/mlserver/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.seldon.ai/mlserver/api-reference/pythonapi.md).

# Python API

MLServer exposes a Python framework to build custom inference runtimes, define request/response types, plug codecs for payload conversion, and emit metrics. This page provides a high-level overview and links to the API docs.

* [MLModel](/mlserver/api-reference/pythonapi/mlmodel.md)
  * Base class to implement custom inference runtimes.
  * Core lifecycle: `load()`, `predict()`, `unload()`.
  * Helpers for encoding/decoding requests and responses.
  * Access to model metadata and settings.
  * Extend this class to implement your own model logic.
* [Types](/mlserver/api-reference/pythonapi/types.md)
  * Data structures and enums for the V2 inference protocol.
  * Includes Pydantic models like `InferenceRequest`, `InferenceResponse`, `RequestInput`, `ResponseOutput`.
  * See model fields (type and default) and JSON Schemas in the docs.
* [Codecs](/mlserver/api-reference/pythonapi/codecs.md)
  * Encode/decode payloads between Open Inference Protocol types and Python types.
  * Base classes: `InputCodec` (inputs/outputs) and `RequestCodec` (requests/responses).
  * Built-ins include codecs such as `NumpyCodec`, `Base64Codec`, `StringCodec`, etc.
* [Metrics](/mlserver/api-reference/pythonapi/metrics.md)
  * Emit and configure metrics within MLServer.
  * Use `log()` to record custom metrics; see server lifecycle hooks and utilities.

{% hint style="info" %}
When creating a custom runtime, start by subclassing `MLModel`, use the structures from [Types](/mlserver/api-reference/pythonapi/types.md) for requests/responses, pick or implement the appropriate [Codecs](/mlserver/api-reference/pythonapi/codecs.md), and optionally emit [Metrics](/mlserver/api-reference/pythonapi/metrics.md) from your model code.
{% endhint %}