Skip to main content

Billing Procedure and Resource Usage

As Clustergate-2 operations move beyond the commissioning and trial phases, this document outlines the billing procedures, resource usage accounting, and how to select a commitment tier that aligns with your operational needs.

Tracked Resources​

Your resource consumption is tracked across the following metrics:

  • Downlink & Uplink (Volume)
  • Compute Runtime (CPU, GPU, and FPGA)
  • Storage (Allocated onboard volume)
  • Dedicated EM Days (Engineering Model access)

Usage Accounting Rules​

Please note the following points regarding how historical and future usage is accounted for:

  • Flight Model (FM) Executions: Historical FM executions are included in your committed usage. Commissioning runs are not included.
  • Engineering Model (EM) Executions:
    • Executions performed on dedicated EM access before launch (December and January) are included in committed usage.
    • Post-launch EM evaluation usage provided for onboarding, testing, and familiarization is not accounted for.
    • As of July 1st, EM access rolls back to an on-demand model. Additional EM capacity can be purchased beyond the committed allocation if needed.

Selecting a Commitment Tier​

To transition to standard billing, please review our options at software.dphispace.com and complete the following steps:

  1. Select a commitment tier.
  2. Specify any additional quota requirements on top of your chosen tier.
  3. Detail any specific operational needs not covered by standard tiers.
  4. Complete the pricing PDF available on the website and return it to our team.

Once your PDF is received, we will review your requirements, issue an invoice, and configure the appropriate quotas on your dashboard upon payment. Note: Your configured quotas will account for your historical FM usage and dedicated pre-launch EM usage.

Onboard Storage Billing​

Storage forms part of your committed resources and is billed based on the amount allocated:

  • A one-time fee is applied during the first month of allocation.
  • A recurring monthly fee applies for persistent storage thereafter.

Exception: Recurring monthly storage fees will not be billed until the platform provides API access to space resources (such as fisheye imagery and satellite telemetry), which is dependent on services provided by Momentus.

Execution Scheduling​

Execution scheduling is performed on a best-effort basis.

  • To optimize spacecraft operations and reduce requested passes, customer executions are preferentially bundled together.
  • During periods of lower overall demand, turnaround times may be longer than during high-activity periods.
  • If you have strict latency or scheduling requirements, please discuss them with our team when selecting your commitment tier.

Metrics and Billing Transparency​

Several dashboard updates have been deployed to improve metrics tracking and billing transparency:

Compute & Executions​

  • Pod Run Duration: The configured max_duration of a pod is accounted for within compute runtime quotas. If a pod fails early due to a user error, the full configured time is still billed. The platform provides an unbilled 10-minute grace period beyond this duration.
  • Docker Builds: Build execution time is included within compute quotas, and generated build logs fall under downlink quotas. If a build fails due to a user-side error (e.g., missing file in a COPY operation), the consumed compute time and log downlink are still accounted for.
  • Pre-Execution Evaluation: Resource requests that can be projected before execution (such as uplink volume and CPU/GPU/FPGA runtime) will be evaluated before scheduling. Requests expected to exceed available quotas may be rejected early. Unpredictable metrics (downlink usage, API calls, generated logs, token consumption) are accounted for after the operation completes.

Downlink & Logs​

  • Log Downlink: Pod run tasks now expose stdout and stderr logs in the user interface. These logs count toward your downlink budget.
  • Listing Operations: File listing and Docker image listing operations count toward downlink quotas.

API & AI Metrics​

  • Telemetry API: Tracked metrics include the number of telemetry API calls, image API calls, and total bytes received from the onboard telemetry service.
  • AI API: Tracked metrics include the number of tokens consumed, total AI runtime, and total bytes received from the onboard AI service.

Usage Reporting​

  • Enhanced reporting is available in the dashboard to provide a clearer breakdown of resource consumption by specific task and execution.