Compare plans
There are two billing plans:
You can’t create dedicated endpoints on the Starter plan. To deploy models on
dedicated infrastructure, you need an Enterprise plan.
Contact us to learn more about
Enterprise plans.
For specific rate limits per tier, see
Rate limits.
Use credits
On the Starter plan, you must have credits to send inference requests. New Modular Cloud accounts include free credits, so you can start making requests right away. Each request deducts credits for input and output tokens. Token prices vary by model deployment—check prices for each model on the Models and Endpoints pages in the console. Prices shown are the cost per 1 million tokens used. Dedicated deployments are priced separately based on the configuration you choose.One credit equals one US dollar (USD). There’s no separate conversion rate for
credits, so your credit balance and all model prices are in the same units.
Add credits
If your credit balance reaches zero, requests that use your API keys return a 401 response. When you need to add credits, follow these steps:- Open the console Billing page.
- In the Payment method section, click Add card.
- Complete the checkout form in Stripe. Stripe then returns you to the Billing page.
- In the Credit balance section, click Buy credits.
- In the Amount (USD) field, enter the amount you want to buy, then click Buy credits.
Monitor billing
The Billing page also lets you monitor billing activity for your organization.- Current billing cycle: monitor the state of the current billing cycle for all models.
- Past invoices: review and export past invoices.
- Transaction history: inspect and export individual transactions.