What it is
Make (formerly Integromat, owned by Celonis) lets you build automations, called scenarios, on a visual canvas. Each app action is a module, and you connect them with routers, filters and error handlers to build branching workflows. It also offers AI agents, an AI assistant called Maia, Make Grid for mapping all your automations, and an MCP server.
Make bills in credits. Most actions use one credit, routers and error handlers use none, and AI features and some advanced modules use a variable number based on tokens, file size or run time. The free plan includes 1,000 credits a month.
Make's own blog puts the Core plan at $9 a month for 10,000 credits on annual billing; the pricing page lists Pro at $16 and Teams at $29 on the same basis, with monthly billing costing more. Unused credits expire at the end of the term.
Our review
Make's case rests on the canvas and the price. Branches, loops and error routes are laid out as a diagram, which makes a 30-module scenario far easier to audit than a long linear list, and at $9 a month for 10,000 credits on annual billing it undercuts Zapier, whose paid plans start at $19.99 for 750 tasks according to Make's own comparison.
The credit model is predictable until AI arrives. A Google Drive upload is one credit whatever the file size, and routers and error handlers are free. Make's AI Provider, the AI Agents and Maia bill on tokens instead; the help centre's own example shows Maia using about 140 credits for a single reply, so a habit of asking the assistant to fix scenarios can eat a Core plan's month in under a hundred messages.
The compliance story has caught up. Make says it has completed SOC 2 Type II and SOC 3 audits, runs an ISO 27001 programme and hosts in US or EU data zones. The EU option has a catch spelled out in its help centre: some AI features and models route to the US or the UK, and bringing your own AI connection removes Make's EU guarantee, so data-residency checks have to be done model by model.
Against n8n the trade-off is plain. n8n charges per workflow run and can be self-hosted for free, so long workflows at volume get cheaper there; Make charges per module and has no self-hosted edition. For operations teams without an engineer to run a server, that is often a price worth paying.
Where it earns its place
An operations manager at a 20-person e-commerce shop on Shopify
Sync orders to accounting, alert on stock levels and route returns
Three scenarios of five to ten modules each running a few hundred times a day fit within Core's 10,000 credits at $9 a month on annual billing; routers and error handlers add nothing.
A marketing ops lead at a 150-person B2B company
Enrich and score inbound leads with an LLM before they reach the CRM
Use a custom OpenAI or Anthropic connection so Make bills one credit per operation and tokens go to your own provider account, rather than Make's AI Provider billing by tokens.
A compliance officer at a German fintech
Approve automations that touch customer data
Put the organisation in an EU data zone and check each AI feature against Make's data-zone table, since some models are served from the US or UK; request the SOC 2 report under NDA.
Before you commit
- Setup
- Browser-based; a first scenario from a template takes minutes, but the data mapping and bundle model take most people some days to learn.
- Integrations
- A large library of app modules, HTTP and webhooks, the Make API on paid plans, an MCP server, Make AI Agents, and enterprise apps such as ServiceNow and Workday.
- Data and privacy
- SOC 2 Type II and SOC 3, ISO 27001 programme, GDPR DPA; US and EU data zones, with some AI features and models routed to the US or UK.
- Cost at scale
- Prices rise with credit volume rather than seats; heavy users of built-in AI should budget by tokens, since a single Maia reply can use over 100 credits.
- Limits
- Credits do not roll over; extra credits expire with the billing cycle, and on annual Core they expire at each monthly reset.
What's good
- Canvas makes branching and error routes easy to audit
- Core from $9/month for 10,000 credits on annual billing
- Routers and error handlers use no credits
- SOC 2 Type II, SOC 3 and ISO 27001; US or EU data zones
What's not
- AI modules and Maia bill by tokens, often well over one credit
- Unused credits expire at the end of the term
- No self-hosted option
- EU data zone does not cover every AI model
Best for
Ops teams and freelancers who build complex visual workflows and care about cost.
Alternatives
Head to head
Make: common questions
- Is Make worth it?
- Make gives the most visual control per dollar in hosted automation, provided someone keeps an eye on what the AI modules are spending. We rate it 4.4 out of 5.
- How much does Make cost?
- Free (1,000 credits/month); Core $9/month, Pro $16/month, Teams $29/month for 10,000 credits billed yearly; monthly billing costs more; Enterprise custom.
- What are the downsides of Make?
- AI modules and Maia bill by tokens, often well over one credit; unused credits expire at the end of the term; no self-hosted option; eU data zone does not cover every AI model.
- Who is Make best for?
- Ops teams and freelancers who build complex visual workflows and care about cost.
- How much does Make cost for a team or in production?
- Prices rise with credit volume rather than seats; heavy users of built-in AI should budget by tokens, since a single Maia reply can use over 100 credits.
- Is Make safe for company data?
- SOC 2 Type II and SOC 3, ISO 27001 programme, GDPR DPA; US and EU data zones, with some AI features and models routed to the US or UK.
- What are the best Make alternatives?
- The closest alternatives we have reviewed are Zapier, n8n and Gumloop.
Answers checked Oct 23, 2026.
Make Make? Show this rating on your site
