> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sqd.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Pricing FAQ

> Frequently asked questions about SQD Cloud pricing.

<AccordionGroup>
  <Accordion title="What's changing with pricing in April 2026?">
    Effective **April 1, 2026**, we're updating our pricing structure:

    * **Cloud Hosting:** 20% increase on all compute resources (API, Processor, Database, Collocated Squids)
    * **Storage:** 20% increase ($0.50 → $0.60 per GB/month)
    * **RPC Add-on:** Overage pricing increases from $2.00 to $5.00 per 1M requests
    * **Free Tier:** Unchanged - 2M RPC requests per month remain free

    **Why the change?** Since launching SQD Cloud over two years ago, we've maintained stable pricing while continuously investing in infrastructure, reliability, and performance. This adjustment allows us to continue delivering high-quality service while investing in new capabilities.

    **When does this take effect?** April 1, 2026. All usage from that date forward will be billed at the new rates.

    For complete details, including pricing tables and FAQs, see the [full pricing announcement](/en/other/pricing-update-april-2026).
  </Accordion>

  <Accordion title="How does billing work? When am I charged?">
    SQD Cloud uses hourly billing for compute resources and monthly billing for storage. You're only charged for the actual time your squids are running.

    **Billing cycle:** Charges are calculated and billed on the first day of each month for the previous month's usage.

    **Prorated charges:** If you deploy a squid mid-month, you only pay for the hours it actually runs. For example, a squid running for 360 hours (half a month) costs exactly half of the monthly rate.

    <Note>
      You can monitor your current usage and estimated costs in the SQD Cloud
      dashboard at any time.
    </Note>
  </Accordion>

  <Accordion title="Why did my cost increase during a replay or zero-downtime update?">
    Every slot is an independent deployment. When an old production slot and a new replay slot run at the same time, both consume billable resources.

    Within each slot:

    * every named processor is billed as a processor service;
    * every API replica is billed separately;
    * Postgres compute and storage are separate;
    * RPC overages are calculated from actual usage.

    Moving a tag changes the stable API destination but does not stop or remove the old slot.

    <Tip>
      Filter [Billing and Usage](https://app.subsquid.io/billing/usage) by squid,
      compare the running slots with your manifest, and remove the old slot after
      the rollback window.
    </Tip>

    See [Slots and tags](/en/cloud/resources/slots-and-tags#zero-downtime-updates) for the cutover procedure.
  </Accordion>

  <Accordion title="How do I identify and reduce the cost of one squid?">
    Start in [Billing and Usage](https://app.subsquid.io/billing/usage) and filter by organization and squid. Match each line with the deployment:

    1. Count running slots.
    2. Count named processors in each slot.
    3. Count API replicas.
    4. Check processor, API, and Postgres profiles.
    5. Check database storage and RPC usage.

    Then optimize the limiting resource:

    * Remove unused slots and tags.
    * Use collocated deployments only for development and testing.
    * Right-size processor and API profiles after reviewing actual utilization.
    * Reduce API replicas only when traffic and availability requirements allow it.
    * Add indexes and query limits before scaling the API or database.
    * Tune [update-heavy tables](/en/cloud/resources/update-heavy-tables) and retention before only increasing storage.
    * Move heavy periodic work to a [background job](/en/cloud/resources/background-jobs) so it does not stall indexing.

    Do not reduce resources solely from a monthly average. Check peak load, replay speed, and production traffic first.
  </Accordion>

  <Accordion title="What is the difference between collocated and dedicated squids?">
    **Collocated squids** share compute resources with other squids on the same infrastructure. This makes them much cheaper (`$0.02/hr` until March 31, 2026, then `$0.024/hr` from April 1, vs separate service costs) but means performance is not guaranteed. If a neighboring squid experiences high load, your squid may become slow or unresponsive.

    **Dedicated squids** have guaranteed compute resources that are never shared. Each service (API, processor, database) runs on its own allocated resources with predictable performance.

    <Warning>
      We strongly recommend dedicated squids for production use. Collocated squids
      are perfect for development, testing, and staging environments.
    </Warning>
  </Accordion>

  <Accordion title="How do I estimate my costs before deploying?">
    Use our [interactive cost calculator](/en/cloud/pricing/overview#estimate-your-costs) to estimate costs based on your configuration. You can also review [real-world examples](/en/cloud/pricing/pricing-details#real-world-cost-examples) for common deployment scenarios.

    <Tip>
      Start with smaller profiles and monitor your actual resource usage. You can
      easily scale up if needed. It's better to start small and grow than to
      overprovision from the start.
    </Tip>
  </Accordion>

  <Accordion title="Can I change my squid's compute profiles after deployment?">
    Yes! You can modify your squid's compute profiles at any time using [manifest-based deployments](/en/cloud/reference/manifest/).

    Simply update your manifest file with the new profiles and redeploy. The new pricing takes effect immediately. See the [profile reference](/en/cloud/reference/scale/) for all available options and their specifications.

    <Info>
      Profile changes require a brief redeployment. Plan accordingly for production
      squids to minimize downtime.
    </Info>
  </Accordion>

  <Accordion title="How is database storage billed? Can I reduce storage costs?">
    Database storage is billed at `$0.50 per GB per month` (until March 31, 2026) or `$0.60 per GB per month` (from April 1, 2026) based on actual usage. This applies to both dedicated and collocated squids.

    **Cost optimization strategies:**

    * Implement data retention policies to remove old data you no longer need
    * Optimize your database schema to reduce redundancy
    * Use appropriate data types (e.g., integers instead of strings where possible)
    * Archive historical data to external storage (S3, BigQuery) if you don't need it for queries
    * Monitor your storage usage in the Cloud dashboard

    <Tip>
      For data-heavy applications, consider a processor-only squid that exports to
      external storage instead of maintaining a large database.
    </Tip>
  </Accordion>

  <Accordion title="What are the 2M free RPC requests? How do I use them?">
    Professional organizations include 2 million RPC requests per month at no additional cost through the [RPC addon](/en/cloud/resources/rpc-proxy). This allows your squids to make blockchain RPC calls without managing your own RPC infrastructure.

    **Overage pricing:** If you exceed 2M requests, additional requests cost `$2.00 per million` (until March 31, 2026) or `$5.00 per million` (from April 1, 2026).

    **How to use:** Configure the RPC proxy in your squid to route requests through SQD's infrastructure. See the [RPC addon documentation](/en/cloud/resources/rpc-proxy/) for setup instructions and rate limiting configuration.
  </Accordion>

  <Accordion title="What happens when I hibernate a squid?">
    When you hibernate a squid, all compute services (API, processor, database) are paused. You only pay for storage costs (`$0.50/GB/month` until March 31, 2026, then `$0.60/GB/month` from April 1) while the squid is hibernated.

    **Benefits:**

    * Preserve your squid's data and configuration
    * Eliminate compute costs when not needed
    * Resume quickly when you need the squid again

    **Perfect for:**

    * Seasonal applications that aren't needed year-round
    * Development squids used only during working hours
    * Backup squids kept ready but not actively running

    <Info>
      Hibernation and resumption are instant operations. You can hibernate and wake
      squids as often as needed.
    </Info>
  </Accordion>

  <Accordion title="How do I decide between collocated and dedicated for my use case?">
    Choose based on your performance requirements and budget:

    **Use collocated when:**

    * Developing or testing new features
    * Running staging environments
    * Building proof-of-concepts or MVPs
    * Budget is the primary concern
    * Occasional slowness is acceptable

    **Use dedicated when:**

    * Running production applications
    * Serving end users who expect consistent performance
    * Processing time-sensitive data
    * You need guaranteed response times
    * Reliability is critical to your application

    <Check>
      Most teams use collocated for development and dedicated for production. This
      gives you cost-effective testing while ensuring production reliability.
    </Check>
  </Accordion>

  <Accordion title="Are there any hidden fees or charges I should know about?">
    No hidden fees! SQD Cloud pricing is transparent and predictable.

    **What you pay for:**

    * Compute (hourly rates for API, processor, database)
    * Storage (monthly rate per GB)
    * RPC requests (only above 2M/month)

    **What's free:**

    * Data egress (outbound traffic)
    * GraphQL API requests (unlimited)
    * SQD Network data access
    * The deployment operation itself; the resources started by a deployment are billed
    * Monitoring and logs

    <Note>
      Variable charges include compute time, storage, and RPC usage above the
      included allowance. Parallel slots and replicas multiply the resources that
      are running.
    </Note>
  </Accordion>

  <Accordion title="Do you offer discounts for high-volume usage or long-term commitments?">
    Yes! We offer enterprise plans with volume discounts, custom SLAs, and dedicated infrastructure for high-volume use cases.

    **Enterprise benefits may include:**

    * Volume-based pricing discounts
    * Custom SLA terms
    * Dedicated support channels
    * Architecture consultation
    * Priority feature requests

    [Book a demo](https://calendly.com/t-tyrie-subsquid/30min) to discuss your requirements and get a personalized quote.
  </Accordion>
</AccordionGroup>


## Related topics

- [Pricing](/en/portal/pricing.md)
- [FAQ](/en/sdk/squid-sdk/substrate/faq.md)
- [Network, Portal and SDK FAQ](/en/other/faq.md)
- [Pricing Overview](/en/cloud/pricing/overview.md)
- [Adding API keys to gateways](/announcements/gateway-api-keys.md)