AI models and credit
Choose a model
Most places Multify runs AI for you let you pick the model:
- In the panel and new tab: click the provider icon in the chat box. The Default row at the top is the model Multify picks for you. Switch when a hard research job needs a stronger model, or a quick rewrite only needs a light one.
- On tasks: when a task is assigned to Multify, set the model that should work it (available on web, desktop, and mobile).
- Favorites: star the models you actually use so the picker stays short and useful.
- Show all / focus: hide the long tail when you only want your team-approved or favorited set.
Prepaid AI credit
If you want to pay for model use directly, you can add prepaid AI credit to your account:
- See your balance in the top navigation while you work.
- Add credit by card when you're ready. If you also have a paid plan, the same card on file is used.
- Optional per-turn cost visibility helps you understand expensive runs without leaving chat.
- We'll email the team when credit runs low so work doesn't stop mid-week by surprise.
Bring your own key (teams)
Team owners can connect their own OpenRouter API keys when they want direct billing with OpenRouter, access to specific model tiers, or independent usage tracking. With a custom key you can also set team defaults for primary, vision, and coding models. If you're not sure you need this, prepaid Multify credit is the simpler path.
Free and paid plans
There are two plans. Free is the full product, with ads. Paid removes the ads and adds a few extras, like your own hosted agent (for example on Telegram) and, on some accounts, scheduled tasks. There's no trial: you're on Free until you upgrade from Billing in the profile menu on app.multify.co. If you reach something that needs the paid plan, Multify tells you and links to the upgrade.
Practical tips
- Use a stronger model for research and multi-step runbooks; use a lighter model for short rewrites.
- Favorite 2–4 models max so the picker stays fast.
- Watch the top-nav balance before a big Monday batch of scheduled work.
- A clear task description helps more than a bigger model.