Rental Types

We're excited to present our flexible rental options tailored to meet diverse computational needs and budgets. Dive into our three primary rental models: On Demand (High Priority), Reserved (High Priority), and Interruptible (Low Priority).

On Demand Rentals:

  • Secure exclusive control over your GPUs with high priority access.

  • Enjoy the freedom to run instances for durations set by the host, visible on the create interface card.

  • Ideal for tasks requiring reliable, continuous operation.

  • Should the maximum duration be reached, the host may renew or conclude the session, with data remaining accessible.

Reserved Rentals:

  • Convert any "on demand" rental into a reserved rental by pre-paying and locking in funds for significant discounts (up to 60%, depending on the provider).

  • Flexibility to withdraw unused funds, adhering to the terms of the discounted rate and duration

  • Offers a balance between commitment and savings, with provisions for fund withdrawal under specific conditions, such as service level agreement failures.

Interruptible Rentals:

  • Operate with a bidding system, where higher bids grant higher priority, offering cost-effective solutions for tasks where execution guarantee is secondary to budget considerations.

  • Perfect for use cases with flexible time frames or those that can handle occasional pauses, offering potential cost reductions of 60% or more.

  • Important to implement data-saving practices, as instances may be paused or stopped based on priority changes or higher priority demands.

Selection Process:

  • Choose between "on-demand" or "interruptible" modes in the create interface, noting that selection is final post-rental.

  • On Demand is suited for critical, uninterrupted workloads, while Interruptible is cost-effective for scalable, less time-sensitive projects.

NodelyAI's rental options are designed to cater to various needs, ensuring that whether you require uninterrupted processing power or are looking to optimize costs with flexible scheduling, there's a solution for you. Explore our offerings to find the perfect fit for your computational tasks.

Last updated