Back to the workbench
02 / LOCAL AI · INFRASTRUCTURE

Home AI.

A private playground for models, tools, and questions that lead to more questions.

Experimental
Local LLMsDockerPythonModel routing

The idea

A place to experiment with local and cloud models, routing, self-hosted interfaces, and tools. The interesting part is finding where a small, private setup can be genuinely useful in everyday workflows.

The first goal is a repeatable baseline: a small set of tasks, a clear record of model settings, and an honest look at response quality and latency.

How the pieces fit

01

Interface

One place to try prompts and compare responses.

02

Route

Explicit model selection and opt-in cloud connections.

03

Evaluate

Repeatable tasks with recorded settings and observations.

From the devlog

Added the experiment outline. First focus: a reproducible local baseline.