Pricing

Understanding Cost in Terms of Computational Energy

At Flux, we've pioneered a unique approach to pricing that mirrors the energy efficiency and resource utilization of our distributed network. Our pricing is not only transparent but also reflects the true computational cost of processing tasks.

The Unit of Measure: Joules

• Why Joules?: In the realm of distributed computing, accurately gauging the computational effort required for machine learning tasks is crucial. Traditional measures like FLOPS (floating-point operations per second) provide a theoretical estimate of processing capabilities but often fail to capture real-world complexities such as data type and operational overheads. At Flux, we use Joules — a direct measure of energy consumption — to quantify the computational resources consumed by a task. This method ensures that you pay for the actual "energy" your tasks utilize, making our pricing model fair and proportional to the workload.

• Calculation of Joules: Each microtask’s energy requirement is calculated based on the complexity of the model and the size of the data segment being processed. This calculation takes into account the type of operations performed and the processing time, providing a precise measure of the energy expended.

Dynamic Pricing in an Open Market

• Market-Driven Costs: Flux operates on a dynamic pricing model similar to an open marketplace. Data providers can bid for processing priority, which means that the cost can vary based on demand and the availability of computational resources in the network.

• How Bidding Works: When submitting tasks, data providers set a bid per Joule, indicating how much they are willing to pay for the processing. Higher bids can lead to faster processing times as they are more attractive to the network of devices, incentivizing quicker task completion.

• Transparency and Control: Our platform provides tools to monitor and adjust bids in real-time, allowing users to manage their costs and processing priorities efficiently. This system ensures that you always have control over how much you spend and how quickly your tasks are processed.

Examples of Pricing Scenarios

• Scenario 1: For a low urgency task, setting a lower bid might mean slower processing times but will reduce costs, suitable for non-time-sensitive tasks.

• Scenario 2: For high-priority tasks, increasing the bid per Joule ensures faster completion, ideal for time-critical operations.

Billing and Credits

• Prepaid Credits: Initially, Flux requires users to purchase credits, which are then consumed based on the Joules used by their tasks. This prepaid model helps manage the risk and ensures that resources are allocated efficiently.

• Post-Usage Billing: As our platform evolves and trust mechanisms solidify, we aim to offer post-usage billing options to provide even greater flexibility for our users.

Getting the Most Out of Your Budget

• Estimating Costs: Before submitting tasks, users can estimate the cost by analyzing the complexity of the task and the current market rates for a Joule. This estimation helps in setting a realistic bid that balances cost and processing time.

• Monitoring and Adjustments: Through our user dashboard, monitor ongoing costs and adjust bids to align with budget constraints and processing needs.

This pricing model is designed to empower you with the flexibility to manage your computational spending effectively while benefiting from the scale and efficiency of our global network. If you need further information or assistance in planning your expenditures on Flux, our customer support team is here to help.