Pricing built for growth

You can deploy open source and custom models on Baseten. Start with an off-the-shelf model from our model library. Or deploy any model using Truss, our open source standard for packaging and serving models built in any framework.

You have control over what GPUs your models use. See our instance type reference for a full list of the GPUs currently available on Baseten. Reach out to us to request additional GPU types.

Yes, new Baseten accounts come with credits so you can get to know the UI and experiment with deployments for free.

Yes, Baseten is SOC 2 Type II certified and HIPAA compliant. You can read more about our SOC 2 Type II certification here. And you can read more about our HIPAA compliance here.

No, you do not pay for idle time – you only pay for the time your model is using compute on Baseten. This includes the time your model is actively deploying, scaling up or down, or making predictions. And you have full control over how your model scales up or down.

Customer support levels vary by plan. We offer email, in-app chat, Slack, and Zoom support. We also offer dedicated forward-deployed engineering support. Reach out to our team to figure out a customer support level that works for your needs.

Yes, discounts on compute can be negotiated as part of our Pro and Enterprise plans. Reach out to our team to learn more.

Yes, you can self-host Baseten in order to manage security and use your own cloud commitments. Talk to our engineers to learn more.