Skip to content
whichdevtool?
Categories ▾

Frameworks and UI

Frontend and full-stack frameworks Backend frameworks UI component libraries

Hosting

VPS and cloud servers App hosting and PaaS

Data

Managed databases File and object storage ORMs and database toolkits

AI

AI model APIs

Email

Transactional email APIs Email marketing and newsletters

Payments

Payments and billing

Domains and edge

Domain registrars CDN, DNS and edge security

Operations and identity

Monitoring, uptime and error tracking Auth and identity Background jobs, queues and workflows
Stacks Directory
Ctrl K

Search directory
Menu
StacksDirectory Pricing changes Glossary
VPS and cloud servers App hosting and PaaS Managed databases Transactional email APIs Email marketing and newsletters Domain registrars CDN, DNS and edge security Monitoring, uptime and error tracking Auth and identity Frontend and full-stack frameworks Backend frameworks UI component libraries AI model APIs File and object storage Payments and billing Background jobs, queues and workflows ORMs and database toolkits
  1. Home
  2. Background jobs, queues and workflows
  3. Hatchet
Ha

Postgres-backed task queue and durable workflow engine · by Hatchet Technologies

Hatchet

MIT-licensed task queue and durable workflow engine built on Postgres, with a cloud offering 1 million free runs a month and per-run pricing after that.

Visit hatchet.run ↗ Alternatives Open source Open source Self-hostable EU region
hatchet.run
Hatchet homepage
§1

Overview

Hatchet is an orchestration engine for background tasks, AI agents and durable workflows. You define tasks as functions in Python, TypeScript, Go or Ruby, run workers that connect to Hatchet, and trigger tasks directly, by event, webhook, cron or schedule. It supports DAG workflows for pipelines and durable tasks that sleep or wait for events and resume after failure.

Unlike Redis- or RabbitMQ-based queues, Hatchet persists every execution in Postgres for the retention period, so failed runs can be inspected and replayed from the dashboard. The maintainers note the trade-off: it has been load-tested to about 10,000 tasks a second but uses more resources than Redis-based systems.

Hatchet Cloud is usage-based. Without a card you get 2,000 runs and events a day; adding a card raises the free allowance to 1 million runs and 1 million events a month, then $30 per million runs and $5 per million events. At a free limit tasks are dropped and new ones error rather than being billed. Custom plans add SSO, RBAC, audit logs, HIPAA and dedicated capacity.

The engine is MIT-licensed and needs only Postgres to self-host, and the cloud data plane runs in the US and Ireland.

§2

Pricing and plans

Free tier · no card

Without a card: 2,000 task runs and 2,000 events a day, 10 workers, 10 cron jobs, 3 users. With a card: 1M runs and 1M events a month free

At the limit Stops at the limit Past a free limit, tasks are dropped and creating new tasks returns errors; you are emailed as you approach the limit.

Plan Price What you get
Pay-as-you-go, no card $0 2,000 runs and 2,000 events a day, 10 workers, 2,000 worker slots, 10,000 backlog
Pay-as-you-go, card on file $0 then usage 1M runs and 1M events/month free, then $30/M runs and $5/M events; 100 workers, 10 GB/day active storage, 50 users
Custom Custom 100M+ runs, dedicated capacity, SSO, RBAC, audit logs, HIPAA BAA, self-hosted or BYOC support
Self-hosted Free MIT licence, Postgres-backed

A run is one task execution; a task that spawns another counts as two, and failed runs count. Active storage above the included 10 GB a day is $0.03/GB-hour; log and trace ingestion above 10 GB is $0.0004/GB-hour; users beyond 50 are $10 each. Billing is monthly in arrears. Data plane regions: AWS Oregon, N. Virginia and Ireland, with Sydney in private beta. Retention is 7 days on pay-as-you-go.

Source: hatchet.run/pricing · prices as published, excluding tax.

§3

What it does

Durable task queue
Execution history is persisted in Postgres, so failed tasks can be inspected and replayed.
DAGs and durable tasks
Build pipelines as DAGs or long-running durable tasks with durable sleep and event waits.
Flow control
Priorities, static and dynamic rate limits, and fair concurrency keyed by user or tenant.
Worker routing
Route tasks by worker labels or weighted affinity, with slot limits per worker.
Observability
Real-time dashboard with alerting, OpenTelemetry collection and Prometheus metrics.
Multi-tenancy
Isolated tenants for dev, staging and prod, with users and roles.
§4

Strengths and limitations

Strengths

  • 1 million free runs a month on cloud once a card is on file
  • MIT licence and only Postgres needed to self-host
  • Failed runs are kept and replayable, unlike most Redis queues
  • EU (Ireland) region available
  • Migration guides from Temporal and Celery

Limitations

  • Without a card the free cloud tier is capped at 2,000 runs a day
  • Tasks are dropped and new ones rejected when a free limit is exceeded
  • Higher resource use than Redis- or RabbitMQ-based queues at very high throughput
  • SSO, RBAC and audit logs are only on custom plans
  • Pay-as-you-go retention is 7 days
§5

Who it suits

Good for

  • Python teams replacing Celery with something observable
  • AI agent and LLM pipelines needing rate limits and fairness
  • Teams that already run Postgres and want to self-host

Look elsewhere if

  • Extreme-throughput queues where Redis-based tools use fewer resources
  • Projects needing SSO on a small budget
§6

Alternatives to Hatchet

Tool From Free option Stages
LH
Laravel Horizon Dashboard and worker manager for Laravel Redis queues
— Open source
Ob
Oban SQL-backed job processing library for Elixir (and Python)
$135/mo Open source
Inngest Durable execution and event-driven workflow platform
$99/mo Free tier
VW
Vercel Workflows Managed durable workflows on Vercel
Usage-based Free tier
SQ
Solid Queue Database-backed Active Job backend for Rails
— Open source
Cloudflare Queues Managed message queue for Cloudflare Workers
Usage-based Free tier

All background jobs, queues and workflows →

§7

Questions

Is Hatchet free?

The engine is MIT-licensed. Hatchet Cloud gives 2,000 runs a day without a card, or 1 million runs and 1 million events a month free with a card on file.

How much does Hatchet Cloud cost?

After the free million, $30 per million task runs and $5 per million events, billed monthly.

What happens when I exceed the Hatchet free tier?

Tasks are dropped and new task creation returns errors; Hatchet emails you as you approach the limit.

What does Hatchet need to self-host?

Postgres, which it uses for both the task runtime and observability data.

Is Hatchet a Celery alternative?

Yes. It supports Python and publishes a Celery migration guide, adding persisted history, DAGs and durable tasks.

Sources read for this page

  • https://hatchet.run/pricing
  • https://github.com/hatchet-dev/hatchet
  • https://raw.githubusercontent.com/hatchet-dev/hatchet/main/LICENSE
Spec sheet
From
Usage-based
$30 per million runs after 1M free (card on file)
Free tier
Without a card: 2,000 task runs and 2,000 events a day, 10 workers, 10 cron jobs, 3 users. With a card: 1M runs and 1M events a month free
Stages
HobbyIndieStartupScale-upEnterprise Suits hobby, indie, startup, scale-up
Does
Task queue, DAG workflows, durable tasks, cron
Free allowance
2k runs/day (no card) or 1M runs/mo (card)
SDKs
Python, TypeScript, Go, Ruby
Self-host
Yes, MIT, needs Postgres
Open source
Yes
Self-host
Yes
EU data
Available
Compliance
SOC 2, HIPAA
Category
Jobs and queues
Visit hatchet.run ↗

To report an out-of-date price or limit, use the correction form.

whichdevtool?

Developer and cloud tools compared on price, free tier and limits, each with its source. For side projects, solo developers and small teams.

Categories

  • VPS and cloud servers
  • App hosting and PaaS
  • Managed databases
  • Transactional email APIs
  • Email marketing and newsletters
  • Domain registrars
  • CDN, DNS and edge security
  • Monitoring, uptime and error tracking
  • Auth and identity
  • Frontend and full-stack frameworks
  • Backend frameworks
  • UI component libraries
  • AI model APIs
  • File and object storage
  • Payments and billing
  • Background jobs, queues and workflows
  • ORMs and database toolkits

Research

  • Directory
  • Stacks by project
  • Developer news
  • Pricing changes
  • Glossary

Site

  • About and methods
  • How we make money
  • Suggest a tool or fix
  • Privacy
© 2026 whichdevtool.com Prices change. Each tool page links to its source. We may earn a commission when you sign up through links on this site. How we make money