Skip to content
Agents tracked: 502 Releases (7d): 613 Agent status: 4 with issuesSince last limit reset: Claude 7d · Codex 5d For AI agents: Markdown Updated Nov 18, 2026

Prefect Cloud

by Prefect

Run and watch scheduled Python jobs.

Integrations and automation

WebsitePricingGitHub

1. Price

Free tier: Hobby plan: 2 users, 5 deployments, 500 min serverless, 7 days retention. As reported on the official pricing page, checked Oct 13, 2026.

prefect.io/pricing

Checked against the maker's official pages, pricing on Oct 13, 2026. Tap a verdict to see its source.How we checkedReport an error

What Prefect Cloud does

Prefect is a free, open-source Python library that turns ordinary Python functions into jobs that run on a schedule, retry when they fail and keep a history of every run. Prefect Cloud is the hosted dashboard that schedules and watches those jobs; you can also run Prefect's own server yourself. It suits data jobs, such as copying and cleaning data every night.

  • Prefect Cloud is one of 17 APIs in the Integrations and automation category on AgentGid.
  • Its cheapest paid plan ($100 a month) is the 13th cheapest of the 15 here with a monthly plan.
  • It is one of 16 that can be started for free.
  • 6 of them ship an official MCP server; Prefect Cloud does not.
  • Its official SDK is the 3rd most downloaded in the category (2.4M a week).

What people build with it

Nightly data copy and cleanup

Pull data from an outside service every night, clean it and load it into your database, with each run logged and visible.

Process many records at once

Run the same step for each customer, file or order in parallel, and see which ones succeeded.

Retry calls to a flaky service

When an outside service fails now and then, retry only that step a few times with a pause, instead of rerunning the whole job.

Slow work behind a web form

Hand long jobs from your web app, such as processing an upload or sending emails, to background workers and show their status.

Wait for a person's approval

Stop a job until someone checks the result and approves it, then carry on.

Weekly report to your team Idea

Build a summary of last week's sign-ups or sales every Monday and post it to your team chat.

Prefect Cloud popularity

2.3M
SDK downloads a week
+10.5% on the 4-week average
18
public agent repos use its SDK
24.2K
GitHub stars, PrefectHQ/prefect
SDK downloads per weekpypi prefect 2,326,081

SDK downloads from npm and PyPI, weekly. Numbers as of Nov 17, 2026.

Pricing

Free tier

Hobby plan: 2 users, 5 deployments, 500 min serverless, 7 days retention

Cheapest plan

$100 a month

Usage price

—

Plans

Added automatically: its SDK was in 8 of the public agent repos we sampled when it was added. Prices read from the official pricing page.

What one action costs

ActionPrice
Starter plan, per month (3 users, 20 deployments, 75 hours of managed compute)$100
Team plan, per user per month (4 to 8 users)$100
Price history
  • A year of the pricing page in the Internet Archive (1 months): the lowest monthly figure went from $100 (2025-11) to $100 (2025-11).
  • 2025-12: free plan shown

As reported on the official pricing page, checked Oct 13, 2026. Prices change: confirm before you buy. prefect.io/pricing

Limits

Hobby (free)

2 users, 5 deployments, 500 minutes of Prefect Serverless a month, 7-day run history

Hobby / Starter / Team

625 / 1,250 / 2,500 API requests and 2,000 / 2,800 / 8,000 logs and events per minute

All Prefect Cloud plans

5 MiB per request body

Prefect Managed runs

4 vCPUs, 16 GB RAM, 128 GB disk, 24 hours per run, monthly compute limit per workspace

From the official docs. prefect.io/pricing

Key and access

How it signs in

API key per user (or service account on higher plans), saved to the local Prefect profile by prefect cloud login, or set as an environment variable with the workspace API URL

Environment variable
PREFECT_API_KEY (with PREFECT_API_URL)
Safe in browser code
No, keep it on your server
Card at sign-up

Not stated on the pages we read

Steps only a person can do

  • Create a Prefect Cloud account (free Hobby plan) and a workspace. Ask the person to run prefect cloud login themselves, or to create an API key for CI.
  • Create an API key with an expiry date and copy it once. Read PREFECT_API_KEY and PREFECT_API_URL from the environment or CI secrets; never commit them.
  • Approve the Prefect connector or plugin if the person wants an AI assistant to read runs. Use its read-only tools; ask before running prefect commands that delete or change deployments.

Commercial use and your data

The Prefect library and server are open source under Apache 2.0, free for paid products. Prefect Cloud is a hosted service under Prefect's terms, and you keep the rights to your data.

In a paid app
Yes

The library and server are Apache 2.0; Prefect Cloud is used under its terms on your own account.

“Apache License Version 2.0, January 2004” source

redistribute-files
With conditions

You may ship Prefect's code under Apache 2.0, keeping the license and notices.

“You must give any other recipients of the Work or Derivative Works a copy of this License” source

Trains on what you send

“Prefect will not, except with Client's prior written consent: (i) train any AI model using Client Data”

Keeps your data

“Prefect Cloud retains flow run, task run, and artifact metadata for a limited time based on your plan.”

Errors to handle

429

Your account sent more requests, logs or events per minute than its plan allows.

The Python package retries up to 5 times using Retry-After; log less, batch work into fewer runs, or upgrade.

413 Content Too Large

One request was bigger than 5 MiB, for example a large artifact or a big batch of logs.

Store large data elsewhere and pass a link, split artifacts, or lower PREFECT_LOGGING_TO_API_BATCH_SIZE.

422

Your prefect package is newer than the self-hosted Prefect server it talks to.

Use the same prefect version on the client and the server.

Crashed

The run stopped because of the machine it ran on, such as running out of memory, a timeout or a killed process.

Check the worker and infrastructure logs, give the run more memory or time, then rerun it.

TimedOut

The run took longer than its time limit.

Raise the timeout, split the work into smaller tasks, or make the slow step faster.

Use it from an agent

Our agent check

4 of 6: llms.txt yes, OpenAPI yes, SDK yes, MCP no, status page no, AI assistants allowed yes

Checked Nov 13, 2026.

MCP server

Prefect MCP server (beta, read-only; also a hosted Claude connector)

claude mcp add prefect -- uvx --from prefect-mcp prefect-mcp-server

PyPI
pip install prefect

Instructions in its docs to check first

The docs say

“curl -LsSf https://astral.sh/uv/install.sh | sh”

Ask the person first. It downloads and runs an installer script for uv on their machine.

The docs say

“Run the following commands in your terminal, after which you'll be prompted in your browser to create or login to your free Prefect Cloud account.”

Account creation and sign-in are the person's steps; do not run uvx prefect-cloud login for them without asking.

The docs say

“/plugin marketplace add prefecthq/prefect-mcp-server”

Ask the person first. It adds a plugin and a connection to their Prefect Cloud workspaces.

Alternatives

Prefect Cloud This page

Free planFree tier

Temporal Cloud

Pick Temporal Cloud for long business processes, such as orders, that must resume exactly where they stopped, in several languages; listed at $0.05 per 1,000 in our catalog on 2026-11-10.

$0.05 / 1k requestsFree tier

Inngest

Pick Inngest for background jobs triggered by events in a JavaScript or Python web app, with a free plan of 50,000 executions a month.

$0.05 / 1k requestsFree tier

Trigger.dev

Pick Trigger.dev when you write TypeScript and want the platform to run long tasks for you, listed at $0.025 per 1,000 in our catalog on 2026-11-10.

$0.025 / 1k requestsFree tier

All integrations and automation APIs

Get started

Tell your AI assistant

“Turn my Python data script into a Prefect flow with prefect pinned to 3.9.2: @task on each step with retries for outside calls, @flow on the entry point, test it locally with prefect server start, then deploy it with a cron schedule; read PREFECT_API_URL and PREFECT_API_KEY from the environment.”

  1. Install Prefect in a virtual environment (Python 3.11 or newer).

    pip install -U prefect

    From github.com/PrefectHQ/prefect/blob/main/README.md

  2. Log in to Prefect Cloud from the terminal through the browser, or paste an API key.

    prefect cloud login

    From prefect.io/docs/v3/how-to-guides/cloud/connect-to-cloud.md

  3. Or skip the account and start a local Prefect server with its dashboard at localhost:4200.

    prefect server start

    From prefect.io/docs/v3/get-started/quickstart.md

First request

01_getting_started.py from the quickstart (comments and blank lines removed), run after uvx prefect-cloud login, so runs show in Prefect Cloud. Prefect is a Python library; Prefect Cloud's REST API is called by the library, not by hand.

from prefect import flow, task
import random
@task
def get_customer_ids() -> list[str]:
    return [f"customer{n}" for n in random.choices(range(100), k=10)]
@task
def process_customer(customer_id: str) -> str:
    return f"Processed {customer_id}"
@flow
def main() -> list[str]:
    customer_ids = get_customer_ids()
    results = process_customer.map(customer_ids)
    return results
if __name__ == "__main__":
    main()

Mistakes to avoid

Running flows with no server set up and relying on the temporary server Prefect starts on its own.
Run prefect server start (or log in to Prefect Cloud) and keep it running while you work.

In our run with prefect 3.9.2 the temporary server stopped when the script ended and printed database errors on the way out, and its history was gone.

Running prefect cloud login in CI or a container without flags.
Pass both --key and --workspace, or set PREFECT_API_KEY and PREFECT_API_URL.

The interactive prompts are not available there and the command fails.

Logging every row or item of a big job.
Log summaries, lower the log level, or batch work into fewer runs.

Each log line counts against the per-minute limit and can trigger 429 answers.

Upgrading prefect in the app but not on the self-hosted server.
Keep the same prefect version on both.

New clients can send fields old servers do not know, which ends in 422 errors.

Leaving usage telemetry on for a client project by default.
Set PREFECT_SERVER_ANALYTICS_ENABLED=false or DO_NOT_TRACK=1 when the person wants nothing sent.

Prefect sends anonymous milestone events and server heartbeats unless turned off.

Questions

Is Prefect free?

The library and the self-hosted server are free and open source. Prefect Cloud has a free Hobby plan; Starter costs $100 a month.

Do I need my own server?

Not to start. Prefect Managed work pools run flows on Prefect's machines with no worker or cloud account; the free plan includes 500 minutes a month.

How long is run history kept?

7 days on Hobby and Starter, 14 days on Team, 30 days on Pro, custom on Enterprise.

Does Prefect train AI models on my data?

No, not without your written consent, according to its terms.

Can my AI assistant look at my runs?

Yes, through the Prefect MCP server, which is read-only and still in beta.