Skip to main content

CelonisWorkflow Automation

Make

Make, formerly Integromat and owned by Celonis, connects 3,500+ apps through visual scenarios built from modules, routers, and iterators. Each module operation is billed in credits. Make AI Agents are in open beta, and the Make MCP server lets assistants such as Claude and ChatGPT run scenarios.

The 30-Second Take

Pick Make if operations teams need routing, loops and error recovery without code, cheaper per step than Zapier.

Teams build scenarios across 3,500+ apps, with routers, iterators and error handlers on one canvas. Celonis owns it, and it now adds AI agents and an MCP server for Claude and ChatGPT.

Best fit
Operations and business teams automating multi-step processes across many SaaS apps.
Main constraint
Nearly every module run uses credits, trigger checks included, and the main admin controls, such as audit logs, SSO, and custom roles, are Enterprise only.
First move
Count module runs for your three busiest scenarios, then pick an annual credit tier and store incomplete executions.

Why

Watch the credits: nearly every module run uses one, trigger checks included, and every public source that discusses them raises them as a concern. Polling schedules and loops set the bill, while routers and error handlers use none.

Practitioner sentiment, all users0 Mixed

48 practitioners, 72 sources reviewed, Updated Nov 1, 2026How this is measured

Read the GILD Review
Save a product your organization uses.Keep up with Make without adding it to your stack.Visit Website
Operating stack
AI Knowledge Work

Comparisons

How does Make compare with the alternatives?

See all 4 comparisons

  1. Make vs Zapier

    • Workflow Automation

    Pick Make for EU data zones and a lower price per step, and Zapier for the larger app catalog. Zapier bills only successful actions, while Make bills a credit for nearly every module run, trigger checks included, and none for routers or error handlers. Count one real workflow on both before comparing prices.

  2. Make vs n8n

    • Workflow Automation

    Pick n8n for long workflows or full self-hosting, and Make for a hosted builder with EU data zones and an on-premise agent on Enterprise. Make charges a credit for nearly every module run, so long scenarios cost more per run. n8n charges one execution per workflow run, whatever its length.

  3. Make vs Power Automate

    • Workflow Automation

    Use the cloud flows your Microsoft 365 licenses include before buying Make, and pick Make when business teams automate many non-Microsoft apps. Make bills credits per module run with unlimited users, in US or EU data zones. Power Automate adds per-user or per-flow licenses and desktop flows for Windows applications.

Operator Insights

What are business and product leaders learning from rolling Make out across their teams?

See all 3 operator accounts

Buying and deployment

What business and product leaders should verify before rolling it out across teams.

Sourced from Make documentation

How you buy it

US$12

Core with 10,000 credits, a month billed annually in the US. Free gives 1,000 and 2 active scenarios.

Source 1, Make, Pricing: PricingSource 5, help.make.com, Help Center full text (Free plan scenario limit): Help Center full text (Free plan scenario limit)

Details

A plan plus a monthly credit tier, priced in US dollars by country. Free: 1,000 credits a month and 2 active scenarios. For US buyers, Core, Pro, and Teams start at 10,000 credits a month for US$12, US$21, and US$38 billed annually, or US$16, US$28, and US$51 billed monthly; buyers in some other countries, such as the UK and Germany, see US$11, US$20, and US$36 annually. Core scales to 300,000 credits a month, Pro and Teams to 8 million, and Enterprise is quoted.

Cost drivers

1 credit per step

Each module operation and trigger check uses one. Extra credits cost 25% more.

Source 2, Make, Credits: CreditsSource 6, Make, Operations: OperationsSource 7, help.make.com, How features use credits: How features use creditsSource 8, Make, Extra credits: Extra credits

Details

Each module operation uses 1 credit for apps without AI, once per bundle it processes, and each trigger check uses 1. Routers and error handlers use none. Make's AI Provider adds credits by input and output tokens, the Code app uses 2 credits per second, and extra credits cost 25% more than plan credits.

Where it runs

Make's cloud

Hosted on AWS in a US or EU data zone, with no self-hosted edition.

Source 9, Make, Security and compliance: Security and complianceSource 3, Make, AI providers and data zones: AI providers and data zonesSource 10, Make, On-premise agent: On-premise agentSource 11, Make, Make MCP server: Make MCP server

Details

Make's hosted cloud on AWS, in the US1, US2, EU1, or EU2 data zone, with no self-hosted edition. Enterprise runs in a separate AWS environment with an on-premise agent for local systems. Claude and ChatGPT reach scenarios through the Make MCP server.

What to verify

Enterprise only

Audit logs, SSO, and team credit limits. Some AI models process in the US whatever the zone.

Source 4, Make, Audit logs: Audit logsSource 12, help.make.com, Single Sign-on: Single Sign-onSource 13, help.make.com, Custom roles: Custom rolesSource 14, help.make.com, Two-factor authentication enforcement: Two-factor authentication enforcementSource 15, Make, Credits per team management: Credits per team managementSource 3, Make, AI providers and data zones: AI providers and data zonesSource 16, Make, Introduction to Make AI Agent (New): Introduction to Make AI Agent (New)

Details

Audit logs, SSO, custom roles, 2FA enforcement, and team credit limits are Enterprise only. Maia and OpenAI models in Make's AI Provider always process in the US, whatever the data zone. Make AI Agent (New) has been in open beta since February 2, 2026, and Make states its functionality and pricing may change.

Signals

What is the current evidence saying about Make?

Public accounts from people running Make at work praise the logic it handles without code and raise credit metering more than any other concern. Every finding links to its sources.

Whose sources count

Practitioner sentiment, all users

0 Mixed net

of 49 sources that take a side, each author counted once.

How this figure is made

Leaders and teams have 19 accounts; GILD prints their sentiment from 20.

Current sentiment

Positive, mixed and negative sources

The score counts them as 14 positive, 20 mixed and 14 negative votes, one per author.

Logic without code draws the most praise, and every source that discusses credits raises them as a concern. Fourteen sources build client scenarios, and 11 describe moving work off Make against 3 that stayed.

Run a pilot on one live workflow to measure failed runs and the time spent fixing them.

Evidence field. All 72 sources, one mark each.

Arrange by

Each mark is one source: positive, mixed, negative, no opinion stated, or documentation (not counted).

By sentiment, All users. Every source GILD read, grouped as the field draws it.
GroupSourcesPositiveMixedNegativeNo opinion stated and not counted in the scoreDocumentation, checked and not counted
Positive15150000
Mixed20020000
Negative14001400
Not counted600060
Documentation17000017
Every source GILD read, by the quarter it was published, as the field's time axis counts it. Documentation is dated by the day it was checked.
QuarterSources
Q1 20256
Q2 20255
Q3 202512
Q4 202511
Q1 202610
Q2 20264
Q3 20267
Q4 20260
Documentation, checked17

By sentiment

17 documentation pages check facts and are not counted.

Hover a mark for a quick look, click it or press Enter to open its source.Tap a mark to open its source, then step to its neighbors with Previous and Next.

What practitioners praise

Top praise of 3. Each mark is a source on it; filled marks praise it.

  1. Logic without code8 of 11 sources on it praise it

Most common concerns

Top concern of 3. Each mark is a source on it; filled marks raise it.

  1. Credit meteringAll 14 sources on it raise it as a concern
All 72 sources behind these findings
  1. Make pricingMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  2. CreditsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  3. OperationsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  4. Extra creditsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  5. Monthly and annual subscriptionsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  6. Adjustments to plans and pricing (Nov 6, 2025)Make (Official Make documentation), Product documentation, checked Oct 12, 2026
  7. Credit usage for AI agentsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  8. Introduction to Make AI Agent (New)Make (Official Make documentation), Product documentation, checked Oct 12, 2026
  9. Make MCP serverMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  10. AI providers and data zones in MakeMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  11. Audit logsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  12. Credits per team managementMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  13. Scenario version controlMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  14. Automatic retry of incomplete executionsMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  15. On-premise agentMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  16. Maia by Make system cardMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  17. Security and complianceMake (Official Make documentation), Product documentation, checked Oct 12, 2026
  18. Automation workflowsMake Community member (self-described: Automation provider with client scenarios), Community forum post, published Sep 28, 2026
  19. How to correctly aggregate and pass 3 parallel AI model outputs into a single Google Gemini judge module in Make?Make Community member (self-described: Self-hosted n8n on own server for a multi-model project), Community forum post, published Sep 28, 2026
  20. n8n always seemed too complicated to me. Until today.Wolfgang Wagner, TYPO3 trainer, integrator, and consultant (self-described: 16 active Make scenarios for his business: RSS imports into Notion and Zoom, Digistore24, and KlickTipp webhooks), Personal blog post, wwagner.net, published Sep 23, 2026
  21. How Stellantis scaled an AI voice agent with MakeDavid Male, Head of Transformation at Stellantis &You UK (Stellantis &You UK's customer service transformation team, scaling an AI voice agent), Make customer story, Stellantis &You UK, published Sep 22, 2026
  22. I deleted the Make.com automation I built for thisR Dilip Kumar, founder building VoiceLab (X profile) (self-described: Make scenario that created his recurring Notion tasks, replaced by a scheduled Claude Code instruction), X post, published Aug 4, 2026
  23. Reporting helpMake Community member, Community forum post, published Jul 29, 2026
  24. Comment on: Idea Farming for MakeAutomation builder, simplymation.eu (self-described: Automation builder with hundreds of production Make scenarios and home automation), Reddit comment, published Jul 2, 2026
  25. Everyone Starts With Zapier. Here's Why I Switched to Make.com Before Paying a CentSEO content specialist (self-described: Content brief automation for their own work), Personal blog post, dev.to, published Jun 27, 2026
  26. Comment on: Tired of Zapier breaking at 2am. What does everyone actually trust for production?Reddit user (self-described: Business automation pipeline on self-hosted n8n), Reddit comment, published May 29, 2026
  27. --blueprint requires inline JSON: Windows argv limit blocks writes of non-trivial scenariosAriel Moatti, developer with an IT specialist background (GitHub profile) (self-described: Windows evaluation of the Make CLI 1.3.1 across 5 Make organizations and 106 active scenarios), GitHub issue on integromat/make-cli, published Apr 23, 2026
  28. Migrate Zapier to Make: A Practitioner's Migration Guide (2026)Automation consultant (self-described: Client content pipelines moved from Zapier to Make over two years), Personal blog post, renezander.com, published Apr 8, 2026
  29. New AI-Agents: feedbackMake user scaling AI agents across scenarios (self-described: AI agents across several scenarios), Community forum post, published Mar 19, 2026
  30. Summry - I replaced my mess of Make.com automations with thisCompetitive intelligence professional (self-described: Company competitive intelligence tracking, about 15 scenarios), Hacker News story text, published Mar 10, 2026
  31. Asana usage running when there's no new activity?New Make user handling client and account management (self-described: Asana task copying for account management at their employer), Community forum post, published Mar 4, 2026
  32. Anyone else getting killed by Make (integromat) operation costs ?Automation freelancer (self-described: Five client accounts), Reddit post, published Feb 26, 2026
  33. Comment on: Anyone else getting killed by Make (integromat) operation costs ?Automation provider (self-described: Client automations), Reddit comment, published Feb 26, 2026
  34. Comment on: Anyone else getting killed by Make (integromat) operation costs ?Reddit user (self-described: Uses Power Automate, n8n, and Azure functions for automation), Reddit comment, published Feb 26, 2026
  35. Comment on: Anyone else getting killed by Make (integromat) operation costs ?Reddit user (self-described: Self-hosted n8n running 30k workflow executions monthly), Reddit comment, published Feb 26, 2026
  36. Comment on: Anyone else getting killed by Make (integromat) operation costs ?Reddit user (self-described: AWS automation setup with 150K+ Make operations), Reddit comment, published Feb 26, 2026
  37. Comment on: Alternative to Zapier? A few observationsReddit user (self-described: Uses Make for Dropbox automation), Reddit comment, published Feb 11, 2026
  38. Show, don't tell: How a Head of Engineering used automation to fuel company-wide transformationChargeGuru Engineering Team, five squads merging legacy systems with Make (ChargeGuru's engineering and product teams, five squads using Make to merge two legacy systems), Make customer story, ChargeGuru, published Jan 23, 2026
  39. Comment on: White House website mysteriously streams personal finance YouTube creatorWebsite operator (self-described: A Make automation that uploads YouTube videos to a website), Hacker News comment, published Dec 20, 2025
  40. 3 automations that made my team think I hired an assistant (it's literally just Make workflows)Reddit user (self-described: Marketing team), Reddit post, published Dec 3, 2025
  41. How we reduced 720K+ Operations Annually Using Make's Custom VariablesMake Community member (self-described: Runs 20 Make scenarios processing monthly client files), Community forum post, published Nov 19, 2025
  42. Comment on: Is Make much worse than Zapier or I'm missing something? Alternatives?Reddit user, Reddit comment, published Nov 17, 2025
  43. Comment on: Is Make much worse than Zapier or I'm missing something? Alternatives?Reddit user, Reddit comment, published Nov 17, 2025
  44. Service Outage: November 12thMake Community member (self-described: Automation provider with client accounts), Community forum post, published Nov 13, 2025
  45. Instagram module suddenly failing with "Media ID is not available (9007, OAuthException)"Make Community member (self-described: Client workflows), Community forum post, published Nov 12, 2025
  46. Trying to use Make with Asana instead of Zapier - Why is there no Trigger on tasks moving to sections yet?Make Community member, Community forum post, published Nov 11, 2025
  47. Triggering scenario in the middle upon users approvalMake Community member, Community forum post, published Oct 27, 2025
  48. Agents-MCP Server with Oauth2?Make Community member (self-described: Team with a self-hosted MCP server), Community forum post, published Oct 26, 2025
  49. Comment on: Introducing Credits: A New System Of BillingMake user, Community forum post, published Oct 3, 2025
  50. FranklinCovey's Make rollout replaced a 15-year-old service busFranklinCovey Business Units, HR, finance, operations, and sales teams building Make scenarios (FranklinCovey's HR, finance, operations, and sales teams, building shared Make scenarios), Make customer story, FranklinCovey, published Sep 30, 2025
  51. How I Saved a Content Agency 80+ Hours/Month with One-Click Content Approvals (Make + Trello)Automation freelancer (self-described: A content agency client with 20 clients), Reddit post, published Sep 29, 2025
  52. Comment on: Introducing Credits: A New System Of BillingMake customer (self-described: Business running scenarios with many Make AI modules), Community forum post, published Aug 29, 2025
  53. Extra Credits Rest?Make Community member, Community forum post, published Aug 28, 2025
  54. Comment on: Introducing Credits: A New System Of BillingMake user who manages two organizations (self-described: Two Make organizations), Community forum post, published Aug 28, 2025
  55. Is it possible to "transfer data"?Make Community member (self-described: Lead generation agency), Community forum post, published Aug 22, 2025
  56. Introducing Credits: A New System Of BillingSolo builder (self-described: Solo builder with a passion project), Community forum post, published Aug 21, 2025
  57. How We Automated Project Administration with Jira, Make and AirtableFounder of Scalevise, a development agency (self-described: Scalevise, internal project administration), Company blog post, dev.to, published Jul 31, 2025
  58. Instagram - Create A Reel For Creator AccountsMake Community member (self-described: Instagram creator account user), Community forum post, published Jul 22, 2025
  59. Comment on: How I build software quicklyRails developer (self-described: Exploring a product idea), Hacker News comment, published Jul 14, 2025
  60. Comment on: What do you expect in return to teach me Make.com?Reddit user (self-described: Automation builder for clients), Reddit comment, published Jul 14, 2025
  61. AI Agents FeedbackAutomation builder for clients (self-described: Client scenarios), Community forum post, published Jul 2, 2025
  62. Make.com AI Agent + Data Store: Key for GetRecordReddit user, Reddit post, published Jun 16, 2025
  63. Hands-Free AI Video: How I Connected Google Sheets and Luma via Make.comIndependent builder (self-described: Personal AI video pipeline), Personal blog post, dev.to, published Jun 2, 2025
  64. How Make helped Stellantis &You UK auto-close 18,000 customer messagesDaniel Cheng, Digital Systems and Database Engineer at Stellantis &You UK (Stellantis &You UK's Aftersales team, analyzing customer messages with Make and OpenAI), Make customer story, Stellantis &You UK, published May 12, 2025
  65. Comment on: N8n - Flexible AI workflow automation for technical teamsHacker News commenter (self-described: Automations and AI agent tasks), Hacker News comment, published May 3, 2025
  66. Diary of a Voice AI Developer 3Freelance voice AI developer (self-described: Freelance low-code automation projects), Hacker News story text, published Apr 14, 2025
  67. Syntaxic recognition of operators, functions and logicals switchMake Community member (self-described: Company evaluating no-code automation), Community forum post, published Mar 28, 2025
  68. My perspective as new user from Make.comMake power user of nearly 8 years who builds automations for clients (self-described: Client automations built inside clients' own Make accounts), Community forum post, published Mar 26, 2025
  69. Where are the AI agents?Make Community member, Community forum post, published Mar 19, 2025
  70. Potentially moving over from Zapier and having difficulty with Squarespace integrationReddit user, Reddit post, published Feb 21, 2025
  71. How can I bill clients based on make usage?Make Community member (self-described: Automation provider billing clients), Community forum post, published Feb 16, 2025
  72. Comment on: How do you use Make in your business?Reddit user (self-described: Freelance automation builder), Reddit comment, published Jan 1, 2025

Jan 2025 to Sep 2026.

55 sources from practitioners, by month and sentiment.

The full readExplore all Make Signals3 praise themes, 3 concerns, 4 points of agreement and 1 disagreement across all three groups, with all 72 sources open to inspect.

Where It Fits

Where does Make belong in the stack?

Open Where It Fits: the Stack Field and the full map

Make is Core across Automate and touches Assist, Find, Create, and Analyze.

Business area
Company-wide
Function
Knowledge Work
Operating stack
AI Knowledge Work
Primary role
Workflow Automation & AI Agents
Make beside the rest of its stackThe Stack Field, every product on every layer
FootprintMake across the AI Knowledge Work stack. Bar height is its role on each layer.

  1. Assist

    1. AI assistant: Touches

      AI assistant, touches. Maia builds and edits scenarios from a chat, and assistants such as Claude and ChatGPT run Make scenarios through the Make MCP server. Open-ended drafting and research stay with a general assistant.

  2. Find

    1. Web research: Touches

      Web research, touches. Make AI Web Search, in beta on all plans, brings live web results into a scenario as structured output. Cited research for people stays with a research tool.

    2. Enterprise search: Touches

      Enterprise search, touches. Knowledge files give a Make AI Agent its own reference documents, embedded for retrieval. Search across the company's apps, with each app's permissions, stays with an enterprise search product.

  3. Create

    1. Documents: Touches

      Documents, touches. Scenarios draft text with AI modules and write it into tools such as Google Docs or Notion, where the team's documents live.

    2. Meeting notes: Touches

      Meeting notes, touches. Make records no meetings. Its scenarios pick up recordings and transcripts from tools such as Zoom and route summaries and action items to other systems.

  4. Analyze

    1. Data analysis: Touches

      Data analysis, touches. Scenarios move and reshape data between systems and keep small datasets in data stores. Analysis and dashboards stay with an analytics tool.

  5. Automate

    1. Automation: Core

      Automation, core. This is Make's primary role. Visual scenarios run across 3,500+ apps with routers, iterators, and error handlers, alongside Make AI Agents in open beta and the Make MCP server. Usage is billed in credits per module operation, on Make's hosted cloud in US or EU data zones.

Market mapThe layers Make covers, and who else owns each one.

  1. Workflow automation & agents

    CoreAutomate

  2. General AI assistant

    TouchesAssist

  3. Web research & cited answers

    TouchesFind

  4. Enterprise search & company knowledge

    TouchesFind

  5. Writing & documents

    TouchesCreate

2 more layers on the full map. Core, Supports and Touches describe coverage. They say nothing about quality.

More on Make

  1. GILD ReviewGILD's evaluation, with the buying facts
  2. Research3 pieces
  3. Guides4 guides
  4. Operator Insights3 operator accounts
  5. Where It FitsOn the AI Knowledge Work stack
  6. Related Products7 products