Inngest
Durable execution for backend and AI workflows, no queues to run.
Overview
Inngest is a durable-execution platform for reliable multi-step functions and AI agent workflows, built by ex-Docker and ex-Buffer engineers since 2021. It runs on serverless, servers or the edge with TypeScript, Python and Go SDKs, and ships AgentKit for multi-agent orchestration.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- 50K executions/month
- 5 concurrent executions
- 500K events ingested
- 1M+ executions/month
- 100+ concurrent executions
- 7-day trace retention
- SAML, RBAC, audit trails
- 90-day trace retention
- Dedicated Slack channel
What you can produce with Inngest
- Durable step functions with automatic retries and state persistence
- TypeScript, Python and Go SDKs
- AgentKit for multi-agent orchestration with MCP tool support
- Runs on serverless, dedicated servers, or the edge
- Built-in tracing, metrics and alerting
- Open-source core engine, self-hostable
ASEAN Perspective
Inngest in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
Inngest solves a genuinely unglamorous problem well: keeping long-running, multi-step jobs and agent workflows alive through crashes, retries and deploys without you standing up a queue and a state store yourself. Developers on Hacker News and in write-ups describe the step-by-step visibility and replay-from-failure as the actual selling point over raw serverless functions. AgentKit, its first-party multi-agent framework, is free and open source and gets singled out for durable, resumable agent networks rather than fire-and-forget calls. The gap people flag: SDK support is strongest in TypeScript and Python, with Go newer and less mature, and the jump from the free Hobby tier (50K executions/month) to Pro at $99/month is steep for a project that's outgrown free but isn't yet enterprise-scale. Backed by $34M in funding including Andreessen Horowitz.
What people say
Inngest's pitch shows up most convincingly in developer write-ups of migrating off raw queue-and-worker setups: people describe testing workflows locally, watching live execution traces, and shipping changes without the usual debugging cycle of tracing a failure back through a message queue. One developer's account, after adopting it for agent workflows, specifically credits the replay feature — rerunning a workflow from the exact step that failed rather than restarting the whole run — as the thing that changed how they thought about reliability.
AgentKit, Inngest's own multi-agent framework layered on top of the durable-execution core, got an enthusiastic reception on Hacker News at launch, with one early user calling the combination with Inngest's realtime features "an awesome combo" despite it being only weeks old at the time. It composes agents into networks with a router and shared state, and uses MCP servers as tools — a fairly opinionated design compared to graph-based frameworks, but one that ships free with the open-source core rather than as a paid add-on.
On the constraint side: SDK coverage is strongest in TypeScript and Python, with Go support existing but newer and thinner, so teams on other stacks are out of luck. Pricing is a real consideration for growing projects — the Hobby plan is free up to 50K executions and 5 concurrent executions a month, generous for prototypes, but Pro jumps to $99/month (some third-party sources cite an earlier $75/month price point, so it's worth checking inngest.com/pricing directly before budgeting) for 1M+ executions and 100+ concurrency. Reviews flag that gap between free and paid as awkward for projects that have outgrown Hobby but aren't ready for Pro-level spend.
The company itself is Series B, $34M-raised, founded in 2021 by Tony Holdstock-Brown (ex-Docker) and Dan Farrelly (ex-Buffer CTO), with Andreessen Horowitz among its backers — reasonably well-capitalized for the category, and the open-source core (5,000+ GitHub stars on the main repo) means the durable-execution engine itself doesn't disappear if the company's roadmap shifts.
Summary of public user & expert reviews, compiled by RECATOOLS.
About this listing
This entry was compiled from publicly available data including Inngest's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Inngest unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Inngest directly →
Spotted something out of date? Suggest an update →
Alternatives to Inngest
More in Agents & Automation
