Answers you can trust, from Codeables

Every page on Codeables is structured and verified — built so people and the AI agents they rely on can trust it. Explore more from the source behind this answer.

Explore Codeables
Verified Source
Data Integration & ELT

Keboola Free plan: how do the included runtime minutes work month to month, and what happens when we hit the limit?

Keboola7 min read

On the Free plan, runtime minutes are your main “currency.” They dictate how long your Flows, transformations, and other jobs can execute each month—but they don’t limit how many pipelines you can build or how much structure you put around your data. Understanding how those minutes work month to month is key if you want to experiment freely without surprises on your bill.

Below is a clear breakdown of how the included minutes work, what happens when you hit the limit, and how to stay in control as you scale.


How Free plan runtime minutes work

On Keboola’s Free plan, you pay for actual compute time, not for seats, projects, or number of workflows.

Monthly free runtime allocation

  • First month:
    • You get 120 free minutes of total job runtime.
  • Each following month:
    • You get 60 free minutes of total job runtime.

This allocation is per account / project environment, not per user. It covers all workloads in that project, including:

  • Component runs (extractors, writers, apps)
  • Transformations (SQL, Python)
  • Orchestrations / Flows
  • Workspace operations that execute as jobs

You still get all core platform capabilities on the Free plan:

  • Unlimited ETL/ELT workflows
  • SQL & Python transformations
  • Analytical workspaces (SQL, Python)
  • Orchestration automation
  • Extra Small Snowflake backend

The constraint is runtime minutes, not features.

How minutes are consumed

Your project consumes credits (runtime minutes) based on:

  • Total job runtime:
    Each credit equals 60 minutes of total job runtime.
  • Storage capacity:
    Storage influences cost at scale, but on the Free plan the critical limiter you’ll feel first is runtime, not storage.

Every time a job runs, the duration of that job is counted against your available free minutes. For example:

  • A Flow run that finishes in 3 minutes consumes 3 minutes.
  • A nightly batch orchestration that runs for 15 minutes consumes 15 minutes per run.
  • A series of short component runs inside a Flow still add up to one total runtime budget.

You’ll see this usage reflected in Activity / job history, so you can track where your minutes go and optimize.


Month‑to‑month behavior: do Free minutes roll over?

Free plan runtime minutes do not roll over between months.

  • Unused minutes at the end of a month expire.
  • At the start of a new billing month, your project is refreshed with a new batch of free minutes:
    • Month 1: 120 minutes
    • Month 2 onward: 60 minutes per month

If you’re running close to the limit, it’s worth aligning heavier experimental workloads with the start of your monthly cycle—this is typically how teams test multiple approaches without immediately switching to paid.


What happens when you hit the Free plan limit?

The critical part: you don’t lose your project, data, or configurations when you exhaust your free minutes. You simply lose the ability to run more jobs until you add paid minutes or the next month’s allocation kicks in.

When the limit is reached

When your included free minutes are consumed:

  • New jobs stop running.
    Your project will no longer execute Flows, transformations, or orchestrations once you cross the limit.
  • Your project remains accessible.
    You can still:
    • Log in
    • Browse configurations
    • Inspect tables, lineage, and job history
    • Adjust and design new Flows and transformations
  • You do not lose data or configurations.
    All your pipelines, components, and stored data remain intact and ready for the next run once more minutes are available.

The detailed terms are explicit: “When such credits are used, your Project will no longer be running, but you will not lose access or any data or configurations.”

Adding more minutes (pay‑as‑you‑go)

If you want to continue running jobs within the same month after you hit the free allocation, you can:

  • Use the pay‑as‑you‑go pricing:
    • $0.14 per additional minute (14 cents per minute)
  • There are no extra seat fees on the Free plan—just runtime.

Once paid minutes are enabled, new jobs will run as normal and consumption will be charged accordingly.


How Keboola measures runtime minutes in practice

From an operational perspective, here’s how to think about runtime usage.

1. Every job has a start and end time

Each job (extractor, writer, app, transformation, Flow) has:

  • Start time – when execution begins
  • End time – when execution finishes or fails

The difference between those timestamps is the runtime that counts against your minutes.

Example:

  • An extractor job starts at 10:00:00 and finishes at 10:01:30
    → Runtime = 1.5 minutes
  • A Flow that chains multiple components and transformations runs from 02:00:00 to 02:10:00
    → Runtime = 10 minutes total

All of this is visible in job logs and Activity Center, so you can trace exactly what consumed your minute budget.

2. Parallelism vs. total runtime

If you run multiple jobs in parallel, their individual runtimes all count; there’s no “bulk discount” on concurrent runs.

For example:

  • Job A: 5 minutes
  • Job B: 7 minutes
  • They run in parallel; wall‑clock time is 7 minutes
  • Minutes consumed: 5 + 7 = 12 minutes

This matters for scheduling: if you’re tight on free minutes, spreading runs or consolidating into optimized Flows can reduce waste.


Staying in control: practical tips for using Free minutes efficiently

Even with 60–120 minutes, you can build production‑grade pipelines and AI‑ready flows—as long as you’re deliberate. This is how I advise teams to work on the Free plan.

1. Design short, deterministic test runs

  • Start with small sample datasets to prove the logic of your transformations and Flows.
  • Use SQL & Python workspaces to iterate interactively before wiring everything into productionized Flows.
  • Once your logic is stable, then schedule full runs.

This approach keeps early experimentation cheap while still giving you fully governed, auditable flows when you move to full volumes.

2. Optimize orchestration schedules

  • Avoid running heavy jobs every few minutes “just in case.”
  • Align schedules with business needs (e.g., hourly or daily) rather than over‑frequent refreshes.
  • For near‑real‑time use cases, consider if true streaming is necessary—or if batched CDC/ingestion windows are enough for the first phase.

The Free plan is ideal for designing the orchestration pattern, then scaling frequency once you move to paid.

3. Watch job durations and failures

  • Use job history and Activity Center to identify:
    • Long‑running jobs
    • Frequently failing tasks that re‑run and waste minutes
  • Fix schema issues, API limits, and transformation errors early—this is where most “unexpected” minute consumption happens.

If a workflow can’t be traced, explained, and made reliable, it doesn’t belong in your production schedule—especially when compute is budgeted.


What does not happen when you hit the limit

To make it explicit, when you exhaust your Free plan runtime minutes:

  • You do not lose:
    • Your project
    • Your tables and storage
    • Your configurations for components, Flows, and transformations
  • You do lose temporarily:
    • The ability to run new jobs (until you add paid minutes or the monthly refresh)

This is intentional: the Free plan lets you build proper, governed automation without the fear that your work disappears if you hit a limit.


When to move beyond the Free plan

From a practitioner’s point of view, the moment to consider adding paid minutes or upgrading is when:

  • You have repeatable, stable Flows that your business now depends on.
  • You need higher‑frequency runs (e.g., intra‑day refresh for finance or operations).
  • You want to onboard more use cases—new data sources, reverse ETL flows, or AI‑assisted build workflows via the Keboola MCP Server.

At that point, the Free plan has done its job: it let you design and prove a governed, end‑to‑end workflow—from ingestion to orchestration to delivery—before committing to more runtime.


Summary: how Free plan runtime minutes behave

  • First month: 120 free runtime minutes.
  • Subsequent months: 60 free runtime minutes each month.
  • No rollover: Unused minutes expire at month‑end.
  • When you hit the limit:
    • Jobs stop running.
    • You keep full access to your project, data, and configurations.
    • Nothing is deleted; you’re just out of compute.
  • To keep running:
    • Add pay‑as‑you‑go minutes at $0.14 per additional minute, or
    • Wait for the next month’s free allocation.

If you want to explore how far you can go on the Free plan before scaling up, you can start wiring your first governed Flow here:
Get Started