Question 1

What is a monthly active user?

Accepted Answer

A monthly active user is a distinct user, device, or other context that your client apps identify to ConfigDirector in a calendar month. An app install or a browser that evaluates configs without identifying anyone counts as one monthly active user too.

Each one counts once a month across all your projects and environments, so the same user ID signed in on a phone and on a laptop counts once. A person who uses your app anonymously and then signs in counts twice: once before signing in and once for their user ID.

Only client SDKs count monthly active users: the browser and mobile SDKs, and the browser side of the Next.js and Nuxt SDKs. Server SDKs are billed in service connection hours instead.

Free and Pay as you go both include 1,000 monthly active users at no cost. The count resets at the start of each billing cycle.

Question 2

What are Service Connection Hours?

Accepted Answer

Service connections are persistent and long-lived connections initiated from server side SDKs. We bill based on service connection hours, meaning the cumulative number of hours server SDKs have been connected or persistently polling.

Free and Pay as you go both include 2,250 service connection hours a month at no cost. One server connected all month uses about 730 hours, plus a buffer for brief scaling up during deploys and traffic spikes, one server is about 750 hours.

For example, the 2,250 free hours cover 3 concurrent servers continuously connected through the month, with a buffer for temporary traffic spikes.

If you configure your server SDK to use polling instead of streaming, each polling request is accounted as 5 minutes of usage for billing. The default polling interval is 5 minutes, matching the same usage rate as a streaming connection.

Question 3

What counts as an Admin API request?

Accepted Answer

Any call to the Admin API counts as one request: from your own scripts, the Terraform provider, the CLI, or your coding agent through the MCP server, where each tool call is one request. Working in the dashboard doesn't count, and neither does evaluating configs in your apps; SDK usage is measured in monthly active users and service connection hours.

The CLI makes two requests for each scan or validate run and three for scan --upload. The ConfigDirector Code Scan action for GitHub makes two when it checks a pull request, and five when you push to your default branch, where it also uploads the scan. A nightly CI job that validates your config keys and uploads the scan makes five requests a night, about 150 a month, well inside the 10K requests free allowance each month.

The MCP server, the CLI, and the agent skills usage is based on the API requests they make, which count toward your usage.

Question 4

What happens when I reach my spending limit?

Accepted Answer

A hard spending limit is optional on Pay as you go. You set it as a dollar amount in the Billing section of the dashboard, and your bill never goes past it in a month.

You won't be caught by surprise. When your usage is on track to go over the limit, you get an alert at 70% and another at 90% of it, and one more if the limit is reached. You can also add up to 5 spend alerts of your own at any dollar amounts, each with a priority from medium to critical.

When spend reaches the limit, service stops until the monthly billing cycle resets: configs are no longer evaluated and each config returns the in-code default value passed to the SDK. The Admin API and the MCP server refuse requests as well. This means your coding agent and the CLI can't read or update configs, and a CI job that runs the CLI fails. Without a hard limit, your flags and API requests keep working whatever your usage, and it is billed at the Pay as you go rates.

Question 5

Is JSON Schema validation included for free?

Accepted Answer

Yes. Both Free and Pay as you go can validate JSON config values against a JSON schema. JSON values are validated against the schema whenever you save the schema itself, update targeting rules, or update experiment variations. A value that does not validate against the schema never reaches your applications.

Validating the provided schema itself, and validating JSON against a schema are two of the most resource intensive operations we run. In order to protect our systems from abuse, Free has an hourly fair-use budget of 20 schema changes and 300 value validations per organization. Each save of targeting rules and experiment variations for a JSON schema-validated config counts as one validation. If you reach the budget, the dashboard tells you when it resets, and your configs keep serving their current values in the meantime.

Pay as you go has no hourly budget on JSON Schema validation. Both options share the same limits on the schema itself: 32 KB per schema, 50 levels of nesting, 1,000 enum items, and 32 regular expression patterns of up to 512 characters each.

Question 6

Can I switch from Free to Pay as you go?

Accepted Answer

Yes. Adding a payment method in the Billing section of the dashboard moves your organization to Pay as you go right away, with no interruption in service. The free allowance still applies, so you pay nothing until your usage goes past it.

If an organization on Free goes over the allowance, it has a 7-day grace period to add a payment method for Pay as you go before service is interrupted. After that, for any usage above the allowance, configs are no longer evaluated and each config returns the in-code default value passed to the SDK until the monthly billing cycle resets.

	Monthly active users	Service connection hours	Admin API requests
Monthly usage
Included free	FreeFirst 1,000	FreeFirst 2,250 hours	FreeFirst 10K
Tier 1	$5.00 / 1K1,000 to 500K	$6.00 / 750 h2,250 to 7,500 hours	$1.00 / 10K10K to 250K
Tier 2	$3.75 / 1K500K to 1M	$5.00 / 750 h7,500 to 22.5K hours	$0.75 / 10K250K to 1M
Beyond	$2.50 / 1KOver 1M	$3.50 / 750 hOver 22.5K hours	$0.50 / 10KOver 1M

Start free. Pay only for what you use.

Pay as you go rates

Estimate your monthly bill

You stay in control of spend

FAQs

Ready to ship confidently?