# What if a small LLM didn't have to make every decision?

> Source: <https://discuss.huggingface.co/t/what-if-a-small-llm-didnt-have-to-make-every-decision/180693#post_1>
> Published: 2026-09-23 11:23:27+00:00

I’ve been experimenting with a different approach to code generation using a small local LLM.

The initial idea was simple: could a 4B model actually generate a usable Python application?

It could.

Well… sometimes.

Instead of trying to make the model responsible for everything, I started moving decisions that can be formalized into a deterministic Python kernel.

**LLM**

**Deterministic kernel**

The result is **Esus**, an open-source experimental code-generation system.

The interesting question for me is not:

*“How do we make a 4B model code like a much larger model?”*

It’s:

**“How much do we actually need the model to decide?”**

I’ve written up the architecture, experiments and benchmark results here:

 [Devops agentique, et si le LLM ne devait pas tout décider ?](https://www.sfeir.dev/devops-agentique-et-si-le-llm-ne-devait-pas-tout-decider/)

The project is open source:

 [GitHub - stengerg-sfeir/esus: Agent Smith · GitHub](https://github.com/stengerg-sfeir/esus)

I’d be interested in feedback from people working with small/local models:

**Which parts of your agents do you think should remain probabilistic, and which parts should be deterministic?**