SQD Cloud Pricing
Use our cost calculator if you need a quick estimate
Effective April 1, 2026, Cloud hosting and RPC pricing will increase. See our pricing update announcement for complete details on the changes.
Free squids
Each account gets a playground organization where a single free squid can be deployed. Certain limitations apply; check the playground organization section for details.
Billing
SQD bills organizations for the resources used by their squids. To deploy production-ready squids and configure their resources, the organization must have a Professional status. All such organizations receive the following perks:
- unmetered GraphQL API requests
- deploying any number of pay-as-you-go dedicated squids
- deploying any number of pay-as-you-go collocated squids
- free 2M RPC addon requests monthly + just
$2/1Mafterwards (increasing to$5/1Mon April 1, 2026) - priority support
Other resources used by squids are priced as described below. Billing typically occurs on the first day of each month.
We do provide enterprise discounts. If your use case requires a lot of resources or an individual SLA, book a demo to get a personal quote.
Pay-as-you-go prices
The pricing is based solely on the resources consumed by the squid, as the SQD Network data is provided free of charge. The total squid price is the sum of compute and database storage prices.
Data egress is not billed.
Storage
Database storage is billed at 0.5$/mo per Gigabyte (until March 31, 2026), both for dedicated and collocated squids. See the pricing update announcement for new pricing effective April 1, 2026.
Compute for dedicated squids
Dedicated squids have separate compute profiles for the Postgres database and API/processor services:
API/Processor
Pricing until March 31, 2026:
small:0.04$/hrmedium:0.08$/hrlarge:0.15$/hrxlarge:0.30$/hr2xlarge:0.60$/hr
API and processor are configured and billed separately. Also, if you use API replicas your cost is multiplied by the number of replicas. For example, running a small processor plus two replicas of medium API costs 0.04+2*0.08=0.20$/hr.
See the pricing update announcement for new pricing effective April 1, 2026.
Full profile descriptions are available here for the API and here for the processor.
Database
Pricing until March 31, 2026:
small:0.08$/hrmedium:0.16$/hrlarge:0.33$/hrxlarge:0.66$/hr2xlarge:1.32$/hr
Profiles are described here. Combine with the cost of storage to get the full price of maintaining the database.
See the pricing update announcement for new pricing effective April 1, 2026.
Hibernated squids
For hibernated squids only the storage costs are charged.
Multi-processor squids
For multi-processor squids each processor is charged separately according to the compute profile.
Collocated (development) squids
Collocated squids share compute resources and thus their performance is NOT guaranteed. We do not recommend using
collocated squids in production, but it's a good fit for development and testing.
Compute is billed at a flat rate of $0.02/hr (~$15/mo) per each collocated squid (API + processor + DB) until March 31, 2026.
It is not possible to additionally configure a collocated squid except for the storage.
See the pricing update announcement for new pricing effective April 1, 2026.
RPC requests
The premium plan already includes a package of 2M requests per month. Above that, it is $2/1M requests (until March 31, 2026). See the pricing update announcement for new pricing effective April 1, 2026.
SLAs
Our SLA only applies to dedicated squids deployed to Professional organizations.
We align with the SLAs of the Google Cloud Platform and will reduce the bill based on the following schedule:
| Uptime | Rebate |
|---|---|
| 95.00% - < 99.50% | 10% |
| 90.00% - < 95.00% | 25% |
| < 90.00% | 100% |
Note that the SLA applies only to the provisioning of the SQD Cloud services (API availability, provisioning of compute resources) and DOES NOT apply to client code. In particular, if a squid is stuck due to a bug or an upstream SQD Network/RPC issue, the SLA discounts don't apply. Similarly, the client is responsible for provisioning compute profiles adequate to the expected traffic.
For Enterprise plan customers, individual SLA terms may be negotiated.
Examples
Small dedicated squid
- 1x API service (small)
- 1x processor (small)
- 1x DB (small)
- 50GB storage
Runs 24/7 (720 hours a month).
| Service | $/hr | $/mo |
|---|---|---|
| API (small) | $0.04 | |
| Processor (small) | $0.04 | |
| Database (small) | $0.08 | |
| Compute total | $0.16 | $115.2 |
| Storage | $25 | |
| Total | $140.2 |
Processor-only squid
A squid writing parquet files to an external s3 bucket. No database or API is provisioned.
- 1x processor (small)
| Service | $/hr | $/mo |
|---|---|---|
| Processor (small) | $0.04 | |
| Compute total | $0.04 | $28.8 |
| Total | $28.8 |
Large production squid
- 2x API service (medium)
- 1x processor (small)
- 1x DB (large)
- 500GB storage
| Service | $/hr | $/mo |
|---|---|---|
| 2xAPI (medium) | $0.16 | |
| Processor (small) | $0.4 | |
| Database (large) | $0.33 | |
| Compute total | $0.53 | $381.6 |
| Storage | $250 | |
| Total | $631.6 |
Test collocated squid:
- API + processor + DB (collocated)
- 50GB storage
| Service | $/hr | $/mo |
|---|---|---|
| Compute total | 0.02 | $15 |
| Storage | $25 | |
| Total | $40 |
FAQ
Q: What is the difference between collocated and dedicated squids? Why are collocated squids much cheaper?
A: Collocated squids share compute resources which means that it may be unresponsive if the neighbour squid is under high load. While we do our best to ensure that the resources are shared fairly, we strongly recommend dedicated services for production use.
Q: I deployed my squid yesterday, will it be billed for the whole month?
A: No, only the actual hours the squid has been running will be billed.
Q: How do I configure the profiles for my squid?
A: Use the manifest-based deployments and the profile reference.
Q: How can I use the provided RPC package? How do I set the rate limits for it?
A: Please look at the RPC addon docs.