Pay Only for What You Execute: Transparent Billing for Modern Workflows

In the world of cloud infrastructure, paying for “seats” feels outdated. When you deploy a serverless application, you don’t pay a monthly fee just to have the account open — you pay for the resources you use and the compute your code actually runs.

Why should your workflow engine be any different?

At mQuark, we believe pricing should align with value. Today, we are introducing the official pricing model for Actionful: a transparent, usage-based architecture designed for developers who demand clarity and efficiency.

It is simple: No platform access fees. No per-seat costs. You pay only for the endpoints you keep active and the workloads you execute.

The Philosophy: Pay for What You Consume

Actionful runs on modern, high-performance cloud infrastructure. Rather than charging a flat rate for “access,” we bill on granular consumption. Whether you are running a lightweight integration or a complex workflow with custom code and external API calls, your costs scale linearly with your actual usage.

The Tiers

Sandbox — Free Forever

Best for: Prototyping, Learning, and Demos.

No credit card. No time limit. Just build.

Cost: $0.00 / month
Rate limits apply to prevent abuse on the shared resource pool
Perfect for proof-of-concept work before you go live

Metered — Pay As You Go

Best for: Startups, Internal Tools, and Live Applications.

When you’re ready to go live, simply add a payment method. There is no monthly platform fee — you pay only for what you execute.

Base fee: $0.00 / month
Billing: Usage-based (see metrics below)
No seat costs, full concurrency, priority compute, and longer log retention

Custom — Dedicated Resources

Best for: Mission-Critical and High-Volume Workloads.

For organizations that need contractual guarantees and dedicated capacity.

Everything in Metered plus dedicated resources
Volume discounts on usage rates
99.9% Uptime SLA and enhanced support

The 5 Billing Metrics

We don’t bundle costs into opaque “credits.” You pay for five specific, measurable activities so you can optimize your architecture with full visibility.

1. Active Endpoints — $5 / endpoint / month

A nominal daily fee per active endpoint covers the reserved capacity needed to serve your live workflows. This is prorated: if you delete an endpoint mid-month, you stop paying for it immediately.

2. API Calls — $0.30 / 1,000 calls

Every time an endpoint is triggered and a workflow executes — whether it succeeds or encounters a rule error — this metric increments. It is the primary measure of your platform activity.

3. HTTP Calls — $0.05 / 1,000 calls

When your workflow calls an external service — a webhook, a REST API, a notification provider — each outbound HTTP request is tracked here. This covers the network and processing cost of your external integrations.

4. Function Invocations — $0.10 / 1,000 invocations

Each time a User Function (your custom JavaScript code) is invoked during a workflow, this metric increments. It measures how often you are using custom compute logic.

5. Function Memory Time — $0.01 / 1,000 GB-seconds

For User Functions, we also measure compute intensity: the duration your code runs multiplied by its allocated memory. This ensures that a quick data transformation and a long-running data-enrichment job are each priced according to their actual resource consumption.

Pricing at a Glance

Metric	Unit	Price
Active Endpoints	per endpoint per month	$5.00
API Calls	per 1,000 calls	$0.30
HTTP Calls	per 1,000 calls	$0.05
Function Invocations	per 1,000 invocations	$0.10
Function Memory Time	per 1,000 GB-seconds	$0.01

Example Monthly Bills

Workload	Estimate
5 endpoints, 10k API calls, 1k function invocations	~$28 / month
2 endpoints, 100k API calls, 10k function invocations	~$41 / month
20 endpoints, 1M API calls, 100k function invocations	~$412 / month

No Hidden Costs

We have designed Actionful to prevent “ghost compute” — the hidden killer of usage-based bills in other platforms.

Actionful enforces strict timeout logic on external HTTP requests. If an external API becomes unresponsive, we cut the connection automatically. You are never charged for idle waiting time.

Failed executions are billed the same as successful ones — compute was consumed regardless of outcome. This is consistent with how all serverless platforms work, and we document it explicitly rather than hiding it in fine print.