Picking the Right Model

Revix ships two AI models, and choosing the right one for the task is the easiest way to get better results while spending less. This guide explains what each model is good at, when to switch, and how model choice affects your usage.

The two models at a glance

Standard	Revix MAX
Best for	Everyday building and editing	Complex systems and hard debugging
Speed	Fast	Slower, more deliberate
Reasoning depth	Light	Deep — thinks through multi-step problems
Usage cost	Low	Higher per message
Availability	All plans	Paid plans

You switch models from the dropdown above the chat input. The selection applies to your next message, so you can change models mid-conversation whenever the task changes.

When Standard is the right choice

Standard is tuned for speed and handles the majority of Roblox development work:

Quick script edits — "make the coin spin faster", "fix this nil error on line 12"
Creating instances — parts, folders, RemoteEvents, lights, spawn points
Property changes — colors, sizes, anchoring, transparency
Simple UI — a shop button, a health bar, a settings panel
Asset searches — finding and inserting Toolbox models
Questions about your game — "where is the damage handled?", "list every script that touches leaderstats"

If a request touches one or two scripts and you can describe it in a sentence, Standard will usually finish it faster than MAX — and cost less.

When to reach for Revix MAX

Revix MAX reasons more deeply before acting, which pays off when the task is genuinely hard:

Multi-file gameplay systems — an inventory with saving, a round-based game loop, a trading system with server validation
Large refactors — restructuring how your RemoteEvents are organized, splitting a monolithic script into modules
Tricky debugging — bugs that span client and server, race conditions, DataStore issues
Anything Standard got wrong twice — if the fast model has failed at a task two times in a row, escalate instead of retrying a third time

MAX takes longer per response and consumes more of your included usage per message, so treat it as the specialist you call in — not the default.

A simple rule of thumb

Start on Standard. Escalate to MAX when the task spans several scripts, needs careful architecture, or Standard has already fumbled it.

This mirrors how professional developers use AI coding tools: fast model for the loop of small edits, strong model for design and diagnosis.

How model choice affects usage

Every message consumes some of your plan's included usage, and Revix MAX messages consume noticeably more than Standard ones. If you burn through your allowance quickly, check whether routine edits are being sent to MAX out of habit — moving everyday work back to Standard stretches the same plan much further. See How Credits Work and Plans Overview for the details of what each plan includes.

Tips for getting the most out of either model

Give the model the context it needs. Name the script or instance you mean ("the ShopHandler script in ServerScriptService") instead of "the shop code". Both models do better with specifics — see Writing Effective Prompts.

One task per message. Three small requests in three messages beats one tangled message, especially on Standard.

Let MAX plan big builds. For large systems, ask MAX to outline the architecture first, then build it. The checklist feature (AI Checklists & Long Tasks) shows you the plan as it executes.

Switch freely. The model dropdown applies per message. Draft the system on MAX, then polish details on Standard in the same chat.