n8n counts "a community of 1.7mn monthly active developers and builders" and more than 1,400 enterprise customers, according to its May 2026 funding announcement, and its GitHub repository had 207,045 stars when we checked on 12 October 2026. That's a lot of people who can build workflows. Knowing how to chain APIs isn't the same as getting paid for it. This guide maps the honest ways to make money with n8n in 2026, checks each one against n8n's own license, and shows where publishing a workflow as an AI agent fits among them.
A correction first. An earlier version of this post said n8n itself doesn't pay builders. That was wrong: n8n runs paid templates for verified creators, a Cloud affiliate program and a partner pilot for agencies, and we've rebuilt the guide around them. For the wider picture, see how to monetize AI agents in 2026.
The honest answer: yes, n8n skills can pay
Yes, and n8n says so itself. Announcing its Series C in October 2025, the company wrote: "We're rapidly seeing people build their own businesses around n8n." The realistic income paths fall into three groups: selling your time to clients, earning through n8n's own programs, and building something that earns without a client brief.
| Income path | How you get paid | Recurring? |
|---|---|---|
| Client workflow builds | A project fee you set | Per project |
| Setup, maintenance, training, consulting | Fees or retainers | Retainers recur |
| Paid templates on n8n's library | One-off sales through your own checkout | No |
| n8n's Cloud affiliate program | 30% of n8n's net earnings on referrals | For 12 months per referral |
| Your own business with n8n behind it | Your customers pay you | If the product sticks |
| Publishing a workflow as an AI agent | A revenue share of every run | While people keep running it |
The fifth path has a showcase on n8n's own site. Its Bordr case study follows a couple who run a relocation-services business on n8n: "What started as a side project turned into a six-figure business in just a handful of months." The skill itself has a tailwind too. In May 2026, SAP made a strategic investment in n8n at a $5.2 billion valuation, with plans to embed n8n in SAP Joule Studio.
The trap is mistaking shipping for selling. You can build a brilliant lead enrichment workflow, post it in the n8n forum, get 400 likes and never see a cent. The likes are real; the income isn't. What closes the gap is a commercial wrapper: someone who pays, a way to collect the money and a reason for buyers to find you.
What n8n itself offers builders now
n8n pays builders directly through one program, its Cloud affiliate commission (30% for 12 months), and lets verified creators sell paid templates through their own checkout. It also runs a closed partner pilot for agencies, but it doesn't process payments for template sellers or pay workflow authors per execution. Its GitHub README now calls n8n a "Fair-code platform to build and deploy AI agents and workflows", and that fair-code license decides what you can charge for (see the license section below).
Paid templates on n8n's library
n8n's template library listed 13,200 workflows on 12 October 2026, with view counts, verified badges and "Buy for $X" buttons on paid listings, so buyer discovery exists. Only verified creators can sell. The n8n Creator hub puts it plainly: "Publish 3 approved templates to become a verified creator." Verified status brings "the ability to publish paid templates".
The money side is yours to run. n8n's template submission guidelines say "you're the one responsible for tracking incoming payments and delivering the template", and warn that "failure to deliver a paid template may result in account suspension or banning." We found no n8n page that mentions n8n taking a commission on template sales, and n8n's docs call the creator program and template marketplace "an ongoing project, and details are likely to change."
We pulled the first 300 templates from n8n's templates API in default order on 12 October 2026. Six were paid, priced $5 to $59 (median $25), and every checkout link went to the seller's own Gumroad (four), Stripe (one) or PayPal (one) page; here's one with a "Buy for $59" button. It isn't a random sample, so don't read it as a share of the library, but the shape is clear: one-off sales from $5 to $59, up against creators like Rahul Joshi, who tops n8n's creators page with 371 templates. The community also runs on a give-first norm that sets a price expectation of zero for most workflows, so the verified-creator route works best as a reputation builder that sometimes sells.
n8n's Cloud affiliate program
n8n's affiliate program pays "30% on all your referrals for 12 months" on n8n Cloud, and its partners page limits that to the Starter and Pro plans. The terms that matter:
- Commission is based on n8n's net earnings for each subscription, and prices are tax-inclusive.
- Payouts go to PayPal monthly, only on balances of EUR 100 or more.
- Paid ad campaigns aren't permitted.
By our rough arithmetic, a Starter referral at 20€ a month billed annually earns you a little under €6 a month for 12 months. With paid ads banned, this path suits people who already have an audience or clients asking which tool to use.
What n8n still leaves to you
If you charge for anything beyond a template, you build the billing yourself: Stripe or Gumroad, usage tracking, support and refunds. That's real work between you and your first sale, on a layer buyers never see. Carrying that layer is what agent marketplaces are for, and where to publish AI agents in 2026 compares the options on exactly this axis.
What n8n's license lets you sell
n8n's license lets you sell your services and the outputs of your workflows. What it restricts is selling n8n itself. The Sustainable Use License says: "You may use or modify the software only for your own internal business purposes or for non-commercial or personal use." n8n's License FAQ, last updated 22 September 2026, turns that into specifics, with one test for grey areas: "is n8n the thing being sold, or is it a tool that supports something else being sold?"
None of the ranking guides for "how to make money with n8n" that we read cite this FAQ, and one recommends selling hosting to clients, which the FAQ rules out. Here's each income path sorted against the FAQ's own words.
| Income path | Status | What n8n's License FAQ says |
|---|---|---|
| Building workflows for clients on your instance | Allowed | "Build automations for your clients on your instance, as long as your clients do not have the possibility to create or edit them." |
| Setup and maintenance fees | Allowed | "Charge your customers for workflow creation, setup, and maintenance services." |
| Training and consulting | Allowed | "Charge your customers for training and consultancy services" |
| Agency work for many clients | Allowed | "There is no limit on the number of clients you can serve from a single instance." The condition: "n8n remains your internal tool, where clients receive outputs, not access." |
| Installing n8n on a client's own server | Allowed with conditions | Fine "as long as you don't offer hosting as well, which would compete with n8n's own Cloud offering." |
| n8n behind your own product | Allowed with conditions | "Use n8n behind the scenes in your own product, where users can use and trigger the automations you built". The license restricts use "where the commercial value of a product derives entirely or substantially from n8n functionality." |
| An AI agent that triggers a workflow you built | Allowed with conditions | "Triggering or executing pre-built workflows via API or webhooks is fine." The grey-area test above still applies. |
| Hosting n8n for clients | Needs a commercial license | Not allowed: "Host n8n as a service and allow your clients to build workflows" |
| Letting end users build their own workflows, including through an agent | Needs a commercial license | Not allowed: "Allow external end users to build or configure their own workflows through your product, whether via a custom UI, our API, MCP, or an AI agent acting on the user's behalf." |
| Forking or white-labelling n8n | Needs a commercial license | Not allowed: "Fork n8n's code to launch your own automation product" and "White-label n8n by removing our branding and offering n8n as is." |
Paid templates and the affiliate program aren't in the table because n8n runs them itself, under the program rules above. Two more notes before you price anything:
- Agents get their own line. An agent that triggers a fixed workflow you built is fine; one that lets its users define workflow logic on your instance is the "AI agent acting on the user's behalf" case the FAQ rules out.
- Grey cases go to n8n. The FAQ points unsure builders to license@n8n.io, and notes that "a partner program covering multi-instance management is in development." Ask before you build a business on an assumption.
Nothing here is legal advice, so read the FAQ yourself before you sign a client. If you publish on Gravity, the builder agreement also asks you to confirm you have the rights to every component your agent uses (clause 7.4), so the license check is yours either way.
Where the paid n8n work is (n8n jobs)
Most paid n8n work is client work, and four places are worth checking first: Upwork, the Jobs category on n8n's forum, n8n's partner pilot and n8n's own careers board. Here's what each showed on 12 October 2026.
Upwork
Upwork's n8n jobs page offered "a sample of the 519 N8N jobs posted on Upwork." Its guide to hiring n8n experts publishes these fixed-price ranges:
| Project type | Upwork's estimated cost |
|---|---|
| Single-workflow automation setup | $500 to $2,000 per project |
| Multi-app integration build | $2,000 to $5,000 per project |
| Complex or high-volume automation pipeline | $3,000 to $8,000 per project |
| Ongoing automation maintenance | $500 to $1,500 per project |
Treat the hourly figure with more care. Upwork says hiring an n8n expert "generally costs $40-$100 per hour, similar to other DevOps engineers", but that range links to its DevOps engineer cost page, and freelancer profiles on the same page listed $8 to $15 an hour. Use the ranges to anchor a quote.
The n8n forum Jobs category
The Jobs category on n8n's community forum held 930 topics all time, including 585 in the past year, 136 in the past month and 19 in the past week, per its Discourse category stats. It mixes "[HIRING]" and "[For Hire]" threads, so it measures activity on both sides of the market, among people who already know what n8n is.
n8n's Experts Partner Program (closed pilot)
n8n's Experts Partner Program is "a closed pilot with founding expert partners" ahead of a full launch. It's recruiting agencies and system integrators that:
- already implement services on top of n8n
- earn their main revenue from automation and AI services
- have at least three customers actively using n8n
- work in the UK and Ireland, Northern Europe or North America
A solo hobbyist won't qualify yet, but the criteria double as a roadmap: three active n8n clients is a sensible first milestone on the way to agency work.
Jobs at n8n itself
If "n8n jobs" means working for n8n, its Ashby job board listed 38 postings, all flagged remote: 20 in Sales, 5 in Product, 4 in Operations, 3 each in Engineering and Marketing, 2 in Finance and Legal, and 1 other. Several are talent-pool entries, so check that a listing is a live opening before you apply.
The bridge: turning an n8n workflow into a marketplace-ready agent
If you want income that doesn't depend on landing clients, publish the workflow as an agent; the bridge is mostly a mapping exercise. In our view, agents that hold up in production share four traits: a single clear job, structured input and output, a quality signal you can measure and an iteration loop. Your n8n workflow probably has the first, and the wrapper supplies the other three.
| n8n workflow element | Agent-ready equivalent | Why the swap matters |
|---|---|---|
| Webhook trigger | LLM-parsed user prompt | Users describe the outcome they want in plain language. |
| Hard-coded credentials | Platform-managed secrets | On a marketplace that carries inference costs, the platform holds the keys and pays for the model, so you don't. |
| Function nodes returning loose JSON | Schema-validated structured output | Downstream consumers and the marketplace validator both need predictable shape. |
| Manual error handling | Retry, fallback, and explain-to-user blocks | End users won't read your error logs; they need a recoverable message. |
| Personal Slack notification | Outcome data + edit signal capture | A quality-ranked marketplace reads these signals; you need them to climb. |
The mapping above is what most "wrap your workflow" tutorials skip. Each row is a small piece of code or configuration. Keep the license line in view while you build: the agent should trigger a workflow you built, with its logic kept out of users' hands (see the license table). How agent tool use actually works goes deeper on the structured output piece.
Step 1: pick a workflow with recurring value (not one-off scripts)
Workflow selection matters more to agent income than engineering polish. In a June 2025 press release, Gartner predicted that "33% of enterprise software applications will include agentic AI by 2028, up from less than 1% in 2024", and also that over 40% of agentic AI projects will be canceled by the end of 2027 "due to escalating costs, unclear business value or inadequate risk controls." Both point one way: narrow, repeatable jobs with obvious value last. Lead enrichment, document parsing and report generation fit; one-off personal automations don't.
Test 1: does someone need this every week?
Run a quick check. Would a real business need this output 4 to 40 times a month, every month, for a year? If the answer's "they'd run it once," you have a script. Scripts belong on GitHub. Agents belong on a marketplace. Recurring value is what makes a per-run revenue share add up to something meaningful.
Test 2: does it generalize beyond your company's data?
Workflows wired into a specific Notion database or one client's Airtable rarely port. Strip the workflow back to the generic version. "Enrich a lead from email" ports; "enrich a lead from email and write it to Aryan's specific Notion column" doesn't. The first 5 AI agents most builders should build lists the workflow shapes that survive this test.
Test 3: is the output measurable?
If a user can look at the output and say "good" or "bad" within 10 seconds, the agent is easier to rank on quality signals. If quality requires three weeks of A/B testing to judge, the marketplace can't surface it. Pick workflows with crisp, fast judgment.
Step 2: wrap it as an agent (LLM-first interface, structured output, clear inputs)
Wrapping is where many builders trip. The wrap layer is small but non-negotiable: an LLM front door, a defined input schema, structured output and clear failure modes. Strict schemas earn their place. When OpenAI introduced Structured Outputs in August 2024, it reported that on its evals of complex JSON schema following, gpt-4o-2024-08-06 with Structured Outputs "scores a perfect 100%", while gpt-4-0613 "scores less than 40%." That measures schema adherence rather than downstream errors, but it shows how much a strict schema changes what comes back.
LLM-first interface
Users don't want to fill a 14-field form. They want to type "enrich these 30 emails with company size and industry" and get a CSV. Put a small LLM in front of your workflow that parses the natural-language ask into the structured inputs your n8n nodes expect. That's one small function call sitting in front of the workflow.
Structured input and output
Define a JSON schema for what comes in and what goes out. Validate both. Reject ambiguous inputs with a helpful error message. Return outputs the marketplace can index, score, and replay. How agent orchestration actually works covers the schema-design piece in detail.
Clear failure modes
If the workflow can't complete, say so plainly. "We couldn't find a company match for 4 of 30 emails, here are the 26 that worked" beats a silent half-result every time. Users forgive partial wins. They don't forgive silence.
Step 3: publish to a marketplace that pays
Publishing is where per-run income becomes real, and what the platform pays for matters as much as the headline split. On Gravity, builders earn 21.27% of every run, up to 25% at the top tier, and Gravity carries the inference and infrastructure costs, so that share is pure profit. Gravity is in private alpha, and builders join through the builder alpha application.
Read the split carefully
A larger percentage where you pay for the model, the hosting, the tool APIs and the support can net less than a smaller share where the platform carries those costs. Which one wins depends on your agent's running costs, so price them before you compare headline numbers. AI agent marketplace splits compared lines the platforms up side by side.
Pick a marketplace with quality-only ranking
If a marketplace sells placement, good agents end up competing with ad budgets. Ranking on outcome data, edit signals and run quality instead lets a narrow expert-built agent beat a well-funded generalist. What makes a top agent on Gravity covers the specific signals.
Read the builder agreement before clicking publish
This is the unglamorous step that protects future you: ownership, takedown rights, payout cadence, refund liability, prompt-injection liability. Two minutes of reading saves months of regret. Gravity's builder agreement is public; give any other platform's terms the same read.
Step 4: iterate based on quality signals (the validator + edit signals + outcome data)
Iteration is what separates the agents that keep earning from the agents that pop and fade. Ship small improvements based on what users actually do with the output, and watch three signal types.
The validator
Expect a quality-ranked marketplace to run an automated test suite against new agents: a battery of prompts, edge cases and adversarial inputs. Pass rate is your first quality signal. Read the failed cases, fix them, ship again. How Gravity runs 80 tests per agent walks through one such suite.
Edit signals
When a user edits the agent's output before using it, that edit is a free training label. High edit volume on a specific field means that field is wrong, vague, or missing. Edits tell you more than ratings do. Ratings are polite; edits are honest.
Outcome data
Did the user actually use the output? Did they re-run the agent next week? Outcome data is the highest-signal ranking input, and re-run rate is the number to watch. An agent people come back to is doing a job they need, and a quality-ranked platform can see that even when buyers can't.
The honest income math
There's no honest single figure for n8n income, so we won't print one. What we can give you is the verified reference point for each path and a formula you fill in with your own numbers. Add up the rows that apply to you.
| Path | Verified reference point | Your monthly formula |
|---|---|---|
| Client builds | Upwork estimates $500 to $2,000 per single-workflow setup | Projects won × your project price |
| Maintenance | Upwork estimates $500 to $1,500 per maintenance project | Retained clients × your fee per maintenance engagement |
| Paid templates | $5 to $59 across the six paid templates in our sample | Sales × price, minus your checkout's fees |
| n8n affiliate | 30% of n8n's net earnings, 12 months per referral, Starter and Pro only | Active referrals × net plan price × 30% |
| Agent on Gravity | 21.27% of every run, up to 25% at the top tier; Gravity carries inference and infrastructure costs | Your share × the revenue your agent's runs bring in |
For the agent row, it helps to know how users pay. Gravity users subscribe to plans that include usage: Basic is free (one agent, no card), Pro is $5 a month during the alpha and Max is $20 a month, and anyone who runs out can buy more usage. The builder's share of every run is pure profit. What moves it is how many people run your agent and how often they come back:
- A niche agent a few people use now and then is pocket money, not rent.
- A solid agent with weekly repeat users is a modest monthly asset.
- A strong agent a team relies on is where a single agent starts to matter.
Income becomes meaningful across a portfolio of narrow agents with repeat users, and only if you keep improving them. The paths also stack: client work can produce a generalized workflow that you sell as a template and publish as an agent, if your client contract allows it. Forget the overnight five-figure screenshots.
What it costs to run n8n
Self-hosting on a DigitalOcean Basic droplet costs $4 a month for 512 MiB of memory, $6 for 1 GiB or $12 for 2 GiB. On n8n Cloud, the Starter plan is 20€ a month billed annually, or 24€ month-to-month, for 2,500 workflow executions, and the Cloud trial needs no credit card. Those are October 2026 prices; our n8n pricing explainer has the full plan table, and the n8n pricing history shows how it has moved. Self-hosting covers your own use and client work on your instance. Hosting n8n for clients is on the FAQ's not-allowed list.
Common n8n income mistakes
These are the six mistakes we'd steer any n8n builder away from. The first five cost you runs and customers; the sixth is a license problem.
- Shipping a personal automation. The workflow that organizes your inbox is a script for you. Generalize it or scrap it.
- Skipping the LLM front door. Buyers expect to describe outcomes in natural language. A bare webhook that only accepts JSON loses everyone who doesn't write JSON.
- Loose output schemas. If the output shape changes between runs, the marketplace can't index it, downstream tools break and edit-signal data becomes noise.
- Burying API costs in the workflow. Hard-coding your personal OpenAI key means you pay every time someone runs the agent. Use the platform's managed credentials or expose the key as a configurable secret.
- Ignoring the iteration loop. Publishing once and walking away is an easy way to stall. We'd aim for a small improvement every week or two for the first quarter.
- Selling what the license doesn't allow. Charging clients to host n8n, white-labelling it or letting customers build their own workflows on your instance all fall outside the Sustainable Use License. Sell services and outputs instead.
FAQ
Is it legal to sell n8n workflows or services?
Yes, within limits. n8n's License FAQ allows charging clients for workflow creation, setup, maintenance, training and consulting while n8n stays your internal tool. Hosting n8n for clients, white-labelling it and letting end users build their own workflows on your instance are not allowed. Grey cases go to license@n8n.io.
Can I sell templates on n8n's template library?
Only as a verified creator, which takes three approved templates. You then collect payment through your own checkout (Gumroad, Stripe or PayPal in our October 2026 sample) and deliver the template yourself, and n8n warns that failing to deliver can get your account suspended or banned.
How much do n8n freelancers charge?
Upwork's cost guide estimates $500 to $2,000 for a single-workflow setup and $2,000 to $5,000 for a multi-app integration build. Its $40 to $100 hourly figure is borrowed from DevOps engineer data, and freelancer profiles on the same page listed $8 to $15 an hour.
Does n8n have an affiliate program?
Yes. n8n pays 30% of its net earnings on referred n8n Cloud subscriptions on the Starter and Pro plans for 12 months. Payouts go through PayPal monthly once your balance reaches EUR 100, and paid ads are not allowed.
Can I keep my workflow on n8n and still earn from it?
Yes, with a license check. n8n's License FAQ says triggering or executing pre-built workflows via API or webhooks is fine, so an agent can call a fixed workflow you built. Letting the agent's users build or configure their own workflows on your instance is not allowed.
Do I need a paid n8n cloud account to monetize a workflow?
No. Self-hosting costs $6 a month on a 1 GiB DigitalOcean droplet, and n8n Cloud Starter is 20€ a month billed annually (24€ month-to-month) for 2,500 executions, as of October 2026. Either way, you may not host n8n for clients.
How is publishing an agent different from selling an n8n template?
A template is a one-off sale: the buyer gets the workflow JSON, you get a fixed price, and on n8n's library you handle payment yourself. A published agent keeps paying for as long as people keep using it; on Gravity that's 21.27% of every run, up to 25% at the top tier.
Do I lose ownership of the workflow when I publish it?
Not on Gravity. Its builder agreement says you retain ownership of the intellectual property in your agent logic, workflows, prompts and configurations, while Gravity gets a non-exclusive license to host, run, evaluate and improve the platform with your agent while it's listed (clause 7.2), plus a reasonable wind-down period after removal.
Sources
We read every page below on 12 October 2026; publication or last-edited dates are given where the source shows one.
- n8n, press release on SAP's strategic investment at a $5.2bn valuation and the Joule Studio plans, PR Newswire, 12 May 2026: n8n valuation and SAP investment press release
- n8n, Series C announcement, 9 October 2025: blog.n8n.io/series-c
- n8n GitHub repository (207,045 stars via the GitHub API) and README: github.com/n8n-io/n8n, README.md
- n8n, Sustainable Use License: LICENSE.md
- n8n, License FAQ, last updated 22 September 2026: n8n License FAQ
- n8n template library and an example paid listing: n8n.io/workflows, n8n.io/workflows/10290
- n8n Creator hub (last edited 27 July 2026) and its child page of template submission guidelines (last edited 10 July 2026): n8n Creator hub
- n8n docs, using templates: Using templates in n8n
- n8n creators directory: n8n.io/creators
- Our template sample: the first 300 results from n8n's public templates API in default order, pulled 12 October 2026 (not a random sample).
- n8n affiliate program and partners page: n8n.io/affiliates, n8n.io/partners
- n8n Experts Partner Program: n8n.io/expert-partners
- n8n, Bordr case study (undated): n8n.io/case-studies/bordr
- n8n Cloud pricing (EUR, annual billing shown by default): n8n.io/pricing
- DigitalOcean droplet pricing: digitalocean.com/pricing/droplets
- Upwork, n8n freelance jobs and the hire-n8n-experts cost guide: upwork.com/freelance-jobs/n8n, upwork.com/hire/n8n-experts
- n8n community forum, Jobs category and Discourse category stats: community.n8n.io/c/jobs, categories.json
- n8n job board on Ashby: jobs.ashbyhq.com/n8n
- Gartner, press release on agentic AI project cancellations and adoption forecasts, 25 June 2025: Gartner agentic AI predictions press release
- OpenAI, announcement of Structured Outputs in the API, 6 August 2024: Introducing Structured Outputs in the API