# Superagent automations default to a faster model

By Steven Van · 2026-09-22

Automations that use the default model now run on GPT 6 Luna at low reasoning, which Base44 says is quicker and uses fewer credits per run.

A [Base44](<https://creatorstoolbox.com/tools/base44>) Superagent automation left on the default model now runs on GPT 6 Luna at low reasoning. Base44 says this is quicker and costs fewer credits per run, as [noted in its changelog](<https://docs.base44.com/changelog/product#september-22-2026>).

Automations where a model was picked by hand are not covered by the change.

[![](<https://docs.base44.com/images/base44-docs-og.png>)](<https://docs.base44.com/changelog/product#september-22-2026>)

![Base44](<https://www.google.com/s2/favicons?domain=base44.com&sz=128>)

Base44

AI app builder — describe what you want and get a fully functional web app in seconds, no coding required.

[View Base44 →](<https://creatorstoolbox.com/tools/base44>)

[Original source](<https://docs.base44.com/changelog/product#september-22-2026>)

[Read Superagent automations default to a faster model on Creators Toolbox](<https://creatorstoolbox.com/blog/base44-superagent-automations-default-to-a-faster-model>)

---
Canonical source: https://creatorstoolbox.com/blog/base44-superagent-automations-default-to-a-faster-model