Zapier vs Make vs n8n: which automation tool should you pick?

The short answer

All three tools connect your apps with “when this happens, do that” workflows, and all three now add AI steps and agents. They differ in who they are built for. Zapier is built for non-technical users who want something working in an afternoon. Make is built for people who think visually and want branching scenarios at a low cost per step. n8n is built for technical teams, or teams with a technical partner, who want long workflows, code when needed and the option to run everything on their own server.

Prices and limits from each vendor’s pricing page and documentation as viewed on September 28, 2026, before taxes.

Current pricing, plan by plan

All three publish entry prices under US$30 a month, but the price you actually pay depends on the volume tier you need. Zapier prices by monthly tasks, Make by monthly credits and n8n by monthly executions, and each discounts annual billing.

Zapier’s Professional plan costs US$19.99 a month on annual billing for 750 tasks (US$29.99 billed monthly), rising to US$49 for 2,000 tasks, US$129 for 10,000, US$289 for 50,000 and US$489 for 100,000. The Team plan starts at US$69 a month on annual billing for 2,000 tasks and 25 users. If you enable pay-per-task, extra tasks cost 1.25 times the base rate on annual plans and 2.5 times on monthly plans.

Make’s Core plan costs US$9 a month on annual billing for 10,000 credits (US$10.59 billed monthly), US$16 for 20,000, US$29 for 40,000, US$55 for 80,000 and US$99 for 150,000. The Pro plan starts at US$16 and Teams at US$29 for 10,000 credits. Make’s November 2025 plan changes priced extra credits 25% above credits included in your plan.

n8n’s cloud Starter plan costs €20 a month on annual billing for 2,500 executions with 5 concurrent executions; Pro is €50 for 10,000 executions and 20 concurrent executions; Business is €667 a month for 40,000 executions and is self-hosted. n8n prices only in euros and advertises 17% savings on annual billing.

Make tier prices read from the credit selector on make.com/pricing on September 28, 2026. n8n Business overage is €4,000 per extra bucket of 300,000 executions.

Tasks vs credits vs executions: why the unit matters

The billing unit matters more than the headline price, because a long workflow multiplies Zapier and Make costs but not n8n costs. A workflow with one trigger and four actions uses four Zapier tasks per run, around five Make credits per run, and one n8n execution per run.

• Zapier: a task is counted whenever Zapier successfully completes a unit of work, including a standard action, an AI by Zapier step or a Code by Zapier step. Triggers, polling, failed actions and built-in tools such as Formatter, Paths and Filter do not use tasks.
• Make: each module action, such as adding a Google Sheets row or fetching Gmail data, counts as one credit. Routers and error handlers do not use credits. Code steps cost 2 credits per second of execution, and AI modules on Make’s own AI provider consume credits based on tokens.
• n8n: an execution is a single run of the entire workflow, regardless of the number of steps or the amount of data processed. Your constraint becomes concurrency (5 simultaneous runs on Starter, 20 on Pro) rather than step count.

Make’s documentation does not state explicitly whether the trigger module of a scenario uses a credit; our worked example below counts it, which is the conservative assumption.

Worked cost comparison at three volumes

For a typical lead-handling workflow with five steps, Make is the cheapest cloud option at every volume, n8n cloud is close and stays flat per run, and Zapier costs several times more as volume grows. The example below is a calculation from published prices, not a client case.

The workflow: a web form submission (trigger) creates a CRM contact, sends a confirmation email, posts a notification to the sales channel and adds a row to a reporting sheet. That is 4 Zapier tasks, 5 Make credits and 1 n8n execution per run. Multi-step Zaps need a paid Zapier plan, and Make’s free plan is limited to two active scenarios and a 15-minute interval, so all three are priced on paid plans.

Self-hosted n8n Community edition has no licence fee at any of these volumes; you pay for the server and for someone to maintain, update and back it up. Add AI usage separately if a step calls a language model.

AI features compared

All three tools can call a language model inside a workflow and all three now offer agent features; the practical difference is how AI usage is billed. Check whether AI steps draw from your normal quota, from a separate AI allowance, or from your own API key billed by the model provider.

• Zapier: AI by Zapier steps count as tasks and are priced by model tier (Standard, Advanced, Premium, or bring your own model). Zapier Agents are billed separately in “activities”, with 400 a month on the free tier. Zapier also sells Chatbots, and Copilot helps build Zaps from a description.
• Make: Maia by Make builds scenarios from a conversation, and every plan includes Make AI Agents (beta), the AI Toolkit and the AI Content Extractor. On Make’s own AI provider, credits follow token usage; for example 3,616 input tokens or 452 output tokens equal one credit on GPT-5 mini models. With your own OpenAI or Claude connection, each AI module operation is one credit and the provider bills the tokens.
• n8n: cloud plans include monthly AI credits (2,300 on Starter, up to 13,700 on Pro), and the Business plan adds an AI Assistant in preview. For self-hosting, n8n publishes a curated AI Starter Kit, which lets you keep AI workflows on your own infrastructure.

Self-hosting and data residency: GDPR and Québec Law 25

If you must control where personal data is processed, n8n is the only one of the three you can self-host, and therefore the only one that can run on servers in Canada or in a specific EU country of your choice. Zapier states that data processed in connection with its services is both transferred to and stored in the US, and it relies on its EU-US Data Privacy Framework certification for European data. Make lets each organization choose a US or EU data center, and that choice cannot be changed after the organization is created. n8n’s cloud plans store data in the EU, on servers in Frankfurt, Germany.

For Québec businesses, Law 25 matters here: before communicating personal information outside Québec, an organization must conduct a privacy impact assessment showing the information will be adequately protected, and the communication must be covered by a written agreement. Every cloud automation tool that handles your clients’ names, emails or phone numbers is such a communication. That does not rule out Zapier or Make, but it means documenting the assessment and the agreement before you connect your CRM.

Self-hosting is not free of work. n8n’s own documentation lists technical expertise as required for installation and configuration of a self-hosted instance, and maintenance becomes your responsibility: updates, backups, security and monitoring. Budget for that time or for a partner who does it.

This is general information, not legal advice. For workflows that make decisions about people, or that process health or financial data, have your setup reviewed by a privacy professional.

Learning curve and day-to-day maintenance

Zapier is the easiest to learn, Make takes a few days to get comfortable with, and n8n rewards people who can read JSON and write a little JavaScript. Whichever you choose, maintenance matters more than setup: workflows break quietly when an app changes a field, a token expires or a form gets a new question.

• Zapier: linear, step-by-step editor with the largest app catalogue and plain-language setup. Branching exists (Paths) but long logic becomes hard to read.
• Make: visual canvas where you see the data flow, with routers, iterators and error handlers. Very readable once you learn how bundles of data move between modules.
• n8n: node-based canvas with code nodes, expressions and sub-workflows. The most flexible, and the one where a technical person saves the most time.
• All three: name workflows clearly, turn on failure alerts, keep a written list of what each automation does and who owns it, and review execution logs weekly for the first month.

When custom code beats all three

Custom code beats Zapier, Make and n8n when the automation is really a product: high volume, complex business rules, strict latency or data requirements, or logic that needs automated tests and version control. At that point, a small service written for your process is often cheaper to run and easier to trust than a very large visual workflow.

• Volume: hundreds of thousands of steps a month push Zapier and Make into their upper tiers, while a custom service costs roughly the same at 10,000 or 1,000,000 runs.
• Logic: pricing engines, scheduling rules, multi-currency invoicing or anything with many exceptions is clearer and safer in tested code.
• Data: you need processing in a specific country, a full audit trail, or no third-party automation platform in the chain at all.
• Reliability: you need retries, idempotency and monitoring you control, not a platform’s best effort.
• Integration depth: the app has an API but no connector, or the connector does not expose the fields you need.

Decision matrix

Use this matrix to shortlist, then prototype your single most important workflow on the top candidate before committing to an annual plan.

What ZeniTech charges

ZeniTech is a web, app, automation and AI agency based in Québec City, working remotely with clients across North America and Europe on Eastern Time, in English or French. We build on Zapier, Make or n8n depending on your volume, budget and data constraints, and we write custom code when none of them fits. Tool subscriptions are opened in your name, and when we write code, the repository is in your name from the first commit.

Conversions use the Bank of Canada rate of 1.4168 CAD per USD (September 28, 2026) and are rounded. Orvel AI is ZeniTech’s agent layer operated in Québec: the orchestration, context and integrations that sit on top of leading language models.

Frequently asked questions

Related

Sources

• Zapier: Pricing (viewed September 2026)
• Make: Pricing (credit selector viewed September 28, 2026)
• Make Help Center: Credits
• Make Help Center: Adjustments to plans and pricing (November 2025)
• Make Help Center: Organizations (data center location)
• n8n: Plans and pricing (viewed September 2026)
• n8n Docs: Choose how to use n8n (cloud vs self-hosted)
• n8n Docs: Host n8n
• Zapier: Data Transfer Impact Assessment
• Commission d’accès à l’information du Québec: main changes under Law 25
• Bank of Canada: Daily exchange rates (September 28, 2026)

Market ranges compiled on September 28, 2026. They are not a ZeniTech quote.