Skip to main content
NodePad treats models as part of the workflow, not a one-time setup choice. A workspace has a default model, and individual branches can use a different model when the task calls for it.

​
Three model decisions

Workspace default

The model new work in this workspace starts from.

Per-branch model

Switch the workspace default before starting a new branch so that branch runs on a different model.

Cross-branch comparison

Run parallel branches with different models so you can compare answers side by side.

​
Why this matters

Different models are good at different things. NodePad lets you act on that without breaking the shape of the work. You can branch a prompt and ask two models the same question. You can keep one branch on a fast model and another on a slower, deeper one. You can preserve a workspace default while routing specific branches to other models. You can bring your own keys when you want the provider relationship to stay with your account.
A clean branching strategy is what makes model comparison useful. If both branches start from the same point, you can compare the answers instead of comparing two unrelated prompt histories.

​
What the picker shows

The exact model list depends on how your NodePad environment is configured. The hosted product, enterprise deployments, and bring-your-own-key setups can expose different providers and model options. If you are self-hosting, the available models depend on the providers and keys your deployment exposes.

​
Continue

Per-branch model selection

Switch models between branches inside the same workspace.

Bring your own keys

Route model requests through your own provider account.

Forking

Branching is what makes model comparison clean.