Latenode
Low-code automation with full code and AI: pay for runtime, not per task.
What is Latenode?
Latenode is a low-code automation platform that connects apps and APIs while letting builders drop into JavaScript and AI models when needed. It bills by execution runtime rather than per task or per operation.
Best known for: Low-code automation priced by execution runtime, not per task
Key features
- Visual low-code workflow builder
- Inline JavaScript and code steps
- Built-in AI and LLM nodes
- Headless browser automation
- Runtime-based pricing
- Self-hosting option
Who Latenode is best for
- Developers who need real code (JavaScript/NPM) inside workflows
- High-volume automations where per-task pricing would be costly
- AI and headless-browser automation in one platform
Ideal team: A technical team running high automation volume that wants runtime-based (not per-task) billing with the option to drop into code.
Latenode pricing
Free
$0
- 10,000 CPU-seconds/mo
- 5 active workflows
- 3-min max execution
Pay As You Go
$0
- No base fee, metered per CPU-second
- Unlimited workflows, 3 workers
- 30-day execution history
Costs to watch before you commit
- CPU-second metering is hard to forecast, long-running browser or AI steps consume more seconds and raise bills.
- Extra workers, longer execution time, larger file sizes, and 1-min polling are paid add-ons, not included.
- The Free tier caps at 5 active workflows, 3-min executions, and 3-day history.
- The pricing model recently changed, older 'Start/Team' flat plans quoted elsewhere may no longer apply.
Latenode pros and cons
Pros
- Billed by compute time rather than per task, so a workflow with twenty steps costs the same as one with two if it runs as fast
- A free tier of 10,000 CPU seconds a month forever, which covers real usage rather than a trial
- Real JavaScript with NPM packages inside workflows, so you are not stuck when the visual builder runs out
- Per-second rates fall as volume rises, from $0.00012 down to $0.00005
- AI agents, RAG, headless browser, and a database included rather than sold as separate products
- 4.9 from 68 reviews, and no credit card required to start
Cons
- 68 reviews is a small evidence base next to Zapier or Make
- CPU-second pricing is harder to forecast than a task count, especially before you have run anything
- A slow or badly written step costs real money, since you are paying for the time it takes
- The free tier caps active workflows at 5 and execution at 3 minutes
- Aimed at technical users; a non-developer will not get the value that justifies choosing it
- Smaller integration library and community than the incumbents
Latenode review: a closer look
What Latenode is
Latenode is a workflow automation platform that connects apps, runs AI agents, drives a headless browser, and lets you drop into real JavaScript wherever the visual builder stops being enough. It competes with Zapier and Make on the same jobs, and differentiates on how it charges: by the compute time your automations consume rather than by how many operations they perform.
Runtime pricing changes which workflows are affordable
This is the whole argument and it is worth understanding properly. Zapier and Make count tasks or operations, so every additional step in a workflow adds cost, and a well-built automation that checks three conditions and updates four records is charged for all seven actions. Latenode charges for CPU seconds. A workflow with twenty steps that completes in a second costs the same as a two-step workflow that completes in a second. The practical effect is that complex automations stop being financially discouraged, and the pattern of splitting logic into fewer, cruder steps to save on task counts disappears. For high-volume or multi-step work the difference is large. For a handful of simple automations a month, the free tier of either platform covers you and this argument is irrelevant.
The flip side of paying for time
The honest counterpart is that inefficiency now has a price. A step that waits on a slow API, a badly written loop, or an AI call that takes eight seconds all bill for the time they consume. Under task-based pricing those cost the same as a fast step; here they do not. That is arguably the correct incentive, since it rewards building things properly, but it does mean cost is harder to predict before you have run anything. Budget by starting on the free 10,000 CPU seconds, measuring what your actual workflows consume, and extrapolating from real numbers rather than estimating up front.
Who this is genuinely for
Latenode is aimed at people comfortable writing code. The ability to use JavaScript and NPM packages inside a workflow is a real advantage for a developer and close to meaningless for a marketer, and the platform's economics reward exactly the kind of dense, multi-step automation a technical person builds. A non-technical team will find Zapier easier, better documented, and connected to more apps, and will not use the capability that makes Latenode cheaper. If nobody on the team writes code, the runtime pricing advantage is unlikely to compensate for the friction.
How it compares
Against Zapier, Zapier has vastly more integrations, better documentation, and a far larger community, at per-task pricing that gets expensive at volume. Against Make, Make is the closest visual competitor and charges by operations, with a more mature interface. Against n8n, n8n is the natural comparison for technical teams and can be self-hosted for infrastructure cost alone, which beats any per-second rate if you are willing to run it. Against Pipedream, that is similarly developer-focused with generous free limits. Against Relay.app, which shut down in September 2026, Latenode is one of the sensible destinations for teams migrating off it. The honest summary: Latenode competes on price model and code access, not on ecosystem.
Who should use Latenode?
A good fit
Latenode fits technical teams and developer-adjacent operators of roughly 1 to 25 running high-volume or genuinely complex automations, where per-task pricing elsewhere has become the constraint on how well things can be built. It suits anyone who has hit the point of simplifying a workflow to save on task counts, and teams wanting AI agents and browser automation without buying them separately.
Look elsewhere if…
Non-technical teams should use Zapier, where the ecosystem and documentation are worth more than the pricing difference. Teams willing to self-host should price n8n, which can be cheaper still. Anyone running a handful of simple automations will stay inside a free tier on any platform and should choose on ease of use. And anyone who needs predictable monthly costs may prefer a task count they can forecast over compute time they cannot.
The verdict
Latenode's runtime pricing is a genuinely different answer to a real problem: per-task billing punishes well-built, multi-step automations, and paying for compute time does not. For a technical team running volume, that can be a large saving, and the included AI agents, browser automation, and real JavaScript make it more capable than the price suggests. Two cautions. Cost is harder to forecast than a task count, so start on the free 10,000 CPU seconds and measure before committing. And the advantage rests on being able to use the code layer, which means a non-technical team is buying the harder tool without the payoff.
Top Latenode alternatives
Popular Automation & iPaaS tools teams evaluate alongside Latenode.
OttoKit (SureTriggers)All-in-one automation connecting WordPress and 1,000+ apps.
Relay.appSHUTTING DOWN: Relay.app closes on 14 September 2026 for paying customers, and new signups are already disabled.
ParabolaTurn messy data into automated workflows, no code required.
ParseurAI-powered data extraction from emails, PDFs, and documents
ActivepiecesOpen-source AI automation and MCP, a self-hostable Zapier alternative.