Hatchet

About Hatchet

Hatchet is an open-source orchestration platform for AI agents, background tasks and workflows that have to survive failure. The core capability is durable execution: a task's progress is checkpointed so a crash, redeploy or timeout resumes rather than restarts, which is the property that separates an agent you can run in production from one you babysit. Around that sit the operational pieces such systems always need and rarely have — cron scheduling, event and webhook triggers, data ingestion pipelines, idempotency keys, batch tasks, throughput controls measured in requests per second, and observability over every run with searchable status, duration and metadata. The dashboard shown on the homepage is the honest tell about who this is for: teams whose problem is not writing the workflow but knowing which of the last hundred thousand runs failed and why. Hatchet is genuinely open source and can be self-hosted, with cloud plans layered on for teams that would rather not run the infrastructure. The cloud pricing model is a base plus usage at $10 per million task runs with the first 100,000 free, which is unusually easy to reason about for this category — most competitors price on compute-seconds or concurrency slots that require a modelling exercise to convert into a bill. SOC 2 Type II is included from the free tier, with HIPAA and audit logs at Scale, and the platform reports processing over 100 million tasks a day.

Key Features

Tags