Hatchet vs Trigger.dev (2026): which should you pick?

Pick Hatchet if

• Fits if you're replacing a queue like Celery with durable workflows
• Works for engineers who want to start free and self-host later

Look elsewhere if

• Skip it if you don't write code, as tasks are defined in an SDK
• Tight if you won't add a card, because the free tier is 2,000 runs a day with 1-day retention
• Watch out: on the free tier, new tasks are dropped past the limit
• Watch out: failed runs count as runs, and SSO and audit logs need Custom

Pick Trigger.dev if

• Fits if you write TypeScript and need jobs that run for minutes or days
• Works for developers building AI agents that must survive crashes

Look elsewhere if

• Not for you if you want a visual builder, as everything is code in your own project
• Tight if you stay free: $5 of credits a month, 10 schedules and 1 day of logs, and tasks stop when the credits are gone
• Pricey if you need more runs, since extra concurrency is $10 a month per 50 and Pro seats beyond 25 are $20 each
• Watch out: compute is billed per second by machine size plus $0.25 per 10,000 runs, so the bill is hard to predict
• Watch out: a HIPAA BAA is a $500 a month add-on

Hatchet

• Without a card on file the free tier is limited to 2,000 task runs and 2,000 events a day, with 1-day retention.
• If a limit is passed on the free tier, new tasks are dropped and creation errors appear.
• Retries are not billed, but failed runs count as runs.
• SSO, RBAC, audit logs and HIPAA BAA are listed only on the Custom plan.

Trigger.dev

• Tasks are written in TypeScript in your own codebase; there is no visual workflow builder.
• Compute is billed per second by machine size, plus $0.25 per 10,000 runs; a Small 1x machine costs $0.0000338 a second.
• A signed HIPAA BAA is a $500 a month add-on on the Hobby and Pro plans.
• The Free plan stops running tasks once its $5 of monthly credits is used up.

Hatchet

Durable execution

Workflows and agents are checkpointed so work resumes after a failure. A failed task can be retried automatically or replayed by hand.

Background tasks and cron

Runs background jobs at scale and creates dynamic cron schedules for recurring workflows.

Webhooks and events

Events and webhooks can trigger workflows, and events can satisfy wait conditions inside a running workflow.

Observability

Traces, logs, metrics and alerts are available for each run, with 7 days of retention on the paid-card pay-as-you-go tier.

Self-hosting

The pricing page says Hatchet is 100% open source under the MIT license, and the Custom plan covers self-hosted and hybrid setups.

Trigger.dev

Tasks with no maximum duration

Tasks are code that run on managed workers with no set time limit, so a single run can last minutes, days or longer. Waiting on a subtask or a long delay does not count toward compute.

AI agents and chat

Agent runs can last minutes or hours and survive crashes. Responses and tool calls can be streamed to a frontend through Realtime, and it works with the Vercel AI SDK.

Schedules and webhooks

Cron-style scheduled tasks and verified webhooks start runs. Schedules are capped by plan, from 10 on Free to 1,000 on Pro.

Concurrency and queues

Queues control how many runs execute at once. Limits only count while code is running, and extra concurrency on Pro is sold in bundles of 50 for $10 a month.

Self-hosting

The platform is open source under the Apache 2.0 license, and the vendor publishes a self-hosting guide for running it on your own infrastructure.

Observability

Runs are monitored and traced in a dashboard, with billing alerts at set share-of-budget levels. Log retention runs from 1 day on Free to 30 days on Pro.

Hatchet

• Pay-as-you-go$0

  to start, then usage-based

  1M runs and 1M events free each month, then $30 per million runs, $5 per million events and $0.10 per GB of durable streams.

• CustomCustom

  For 100M+ runs a month: dedicated capacity, SSO, RBAC, audit logs, HIPAA, SLAs, and self-hosted or BYOC support.

Trigger.dev

• Free$0

  $5 monthly credits, 20 concurrent runs, 5 team members, 10 schedules and 1 day of log retention.

• Hobby$10

  per month, with $10 of usage credits included

  50 concurrent runs, 5 team members, 100 schedules, 7 days of log retention and 3 alert destinations.

• Pro$50

  per month, with $50 of usage credits included

  200+ concurrent runs, 25+ team members ($20 per extra seat), 1,000+ schedules, 30 days of logs and Slack support.

• EnterpriseCustom

  Everything in Pro plus custom log retention, priority support, role-based access control, SSO and SOC 2 and pen-test reports.