AI settings & models
The MadMax default model vs bringing your own: how model selection works for the chat and for AI steps, and how AI spend is metered.
Two different "which model?" questions live in MadMax — the chat that builds, and the AI steps inside your workflows. Both default to the same answer.
The MadMax default
Out of the box, everything runs on the MadMax platform model — picked and tuned by us, billed through your workspace's AI credits, no keys required. When an agent step's card says MadMax default, that's what executes.
Bringing your own model (BYO)
Connect an AI provider (Anthropic, OpenAI, Gemini, …) under Settings → connectors and you can:
- Pick it for the chat — the model chip in the composer lists your connected providers; your choice is remembered per user.
- Bind it to an agent step — the step's connection field plus
its model picker. Two rules worth knowing:
- The model field is honored only with a connection bound — unbound steps always run the platform default (the card says so).
- BYO calls bill to your provider account, not MadMax credits.
Spend
Platform AI usage meters per call into your workspace ledger — the admin section shows usage; builds show a checkpoint if a single build runs unusually hot, asking before it continues. BYO usage is invisible to our meter by design (your key, your bill).
FAQ
Which is better? The default is tuned for building (fast, cheap, structured-output-safe). BYO makes sense when you have credits elsewhere, a compliance need, or a model preference for a specific agent step.
Do web-search-enabled agent steps work on BYO? The live web-search tool is a platform-model feature; BYO connections ignore the toggle (the field's hint says so).