Head to head · Workflow automation · October 2026 research run

Make API + MCP vs Windmill API + MCP

Make API + MCP scores 64.6 (B) on agent readiness against Windmill API + MCP's 48 (D), and leads in 2 of 7 scored categories. Windmill API + MCP leads on schema & documentation, agent ergonomics, payments & pricing and maintenance & community. Both do workflow automation.

Best workflow automation platforms with APIs for AI agents · All 174 workflows comparisons

Which one, for what

Make API + MCP B

Best for Teams who build in a visual editor and want an agent to run those scenarios as tools, cloud only.

Ahead on

  • Reliability, 73 against 15
  • Security & auth, 68 against 60

Also in its favour

  • Free to start without a card
  • No incidents deducted, where Windmill API + MCP loses 11 points for them

Watch for

No published OpenAPI file and no official SDK for the REST API

Windmill API + MCP D

Best for Engineering teams who'd rather write steps in Python, TypeScript, Go or SQL and expose them as tools.

Ahead on

  • Schema & documentation, 79 against 66
  • Agent ergonomics, 73 against 63
  • Payments & pricing, 53 against 45
  • Maintenance & community, 87 against 56

Also in its favour

  • Open source
  • Rated higher by the reviewer agents, 3.0 against 2.5 out of 5

Watch for

The default MCP URL puts the token in ?token= unless a superadmin turns that off

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Reliability Make API + MCP +58

Make API + MCP 73

Statuspage at status.make.com with incidents per zone (20). history.atom, re-read on 4 November, lists ten incidents from 6 August, each one zone, one app or a subset of scenarios. They are AI models on eu1 and us2 (10 August), a subset of scenarios disabled after a release and re-enabled after a rollback (14 August), saving and manual runs on EU2 (24 August), delays on EU1 for 20 minutes (26 August) and US2 for 45 minutes (2 September), degradation on eu1.make.com (9 and 10 September) and us1.make.celonis.com (18 September), UI logs with the API unaffected (23 September), scenarios using the Scenario app (12 October) and shared-webhook events not processed for 27 minutes (2 November). All partial, so minor only (20). The 1 October note read no August or September entries; the feed lists them now. API limits published per organisation, 60 calls a minute on Core, 120 on Pro, 240 on Teams and 1,000 on Enterprise (15). The 429 answer says "please try again later", with no Retry-After header, backoff or safe-retry advice for writes (8). Celonis's legal page for Make lists a "Support Service Description and Service Level Agreement for Make" for Enterprise, on a document host whose robots.txt disallows every path, so its terms are unread. An SLA whose terms we couldn't read isn't credited, as serpkite, socure-riskos, affinity and cloudflare-workers-ai were read (0). The API and the hosted MCP server aren't marked beta (10). 20+20+15+8+0+10 = 73.

Windmill API + MCP 15

Read with the hosted lines. status.windmill.dev redirects to an UptimeRobot page. The page draws its services and history from stats.uptimerobot.com/api/, which that host's robots.txt closes, and the HTML read on 4 November shows only an error fetching the data, so no component or history was read and the status-page line has no evidence (0). No readable history (5, our limit rather than a finding against Windmill). No API request limit or 429 guidance published. Concurrency limits exist but you set them per script (0 and 0). The Enterprise plan lists SLA and priority support with a 3-hour response time, and no uptime SLA document was found, so the listed SLA has nothing behind it to read (0). The API and MCP server are generally available (10).

Maintenance & community Windmill API + MCP +31

Make API + MCP 56

Make's official plugin and skills repository (integromat/make-skills, owner Make, support@make.com) is a dated vendor source. Plugin version 0.2.0 shipped on 6 October 2026 with skills rewritten for the v2 MCP server, 29 days before this check (30). New plugin directories for Devin and GitHub Copilot (7 October), Grok (8 October) and OpenClaw (22 October) carry the same 0.2.0 version and the same skills, synced from the Codex plugin, so they are packaging for further marketplaces, not new releases. The MCP server was published to the registry on 7 October and moved to com.make/mcp-server 2.0.0 on 30 October, a listing change recorded but not counted. The repository has no tags, its CHANGELOG stops at 0.1.7 (29 May), and the white-label release notes still stop at 2026.05, undated. One dated release in the 90 days to 4 November (0). Public release notes with dated deprecations and a community forum. We didn't test a support channel (8). Listed in the official MCP registry as com.make/mcp-server, per the repository's server.json (15). No official SDK for the REST API, and the legacy npm MCP server has had no release since v0.5.0 (3). 30+0+8+15+3 = 56.

Windmill API + MCP 87

v1.860.0 released on 3 November 2026, 1 day before this check (30). 100 dated changelog releases from 1.781.1 (6 August) to 1.860.0 (20). Open issues per the 1 October check, not re-read on 4 November, 569 open, most of the ten newest unlabelled (12). Official clients on npm (windmill-client) and PyPI (wmill) published by release workflows in the repository (15). CI runs backend tests on Linux and Windows, frontend checks, CLI and SDK tests (10).

Schema & documentation Windmill API + MCP +13

Make API + MCP 66

No OpenAPI file. MCP tools get typed inputs from each scenario's input definitions, so half credit (10). llms.txt with about 380 entries and a Markdown copy of every page (10). API reference pages state each endpoint's purpose. MCP tool descriptions come from whatever the scenario owner wrote (10). Typed, documented parameters on the API, and scenario inputs typed by their owner (10). A single error schema with detail, message and code, Make-specific codes (IM001 to IM102, SC403) and request examples (13). Versioned at /api/v2, with white-label release notes and dated deprecation notices (13). Re-read on 4 November, llms.txt has 541 lines, and its pages carry a GitBook "Agent Instructions" block, treated as data. 10+10+10+10+13+13 = 66.

Windmill API + MCP 79

OpenAPI 3.0.3 file in the repository with 947 operations on 4 November, versioned with each release (1.860.0) and Apache-2.0 licensed. Every script and flow gets a JSON Schema from its signature (25). llms.txt, read on 4 November, and searchDocs and readDocsPage hand the model Markdown docs from inside the MCP server (10). The 42 endpoint tools reuse OpenAPI summaries. The docs tools say when to call them, most others only say what they do (14). Typed parameters with required fields, but flow bodies are OpenFlow objects of nested free-form values (11). Per the 1 October count, 846 response entries are 200s and only about 30 document an error code (6). A release-please CHANGELOG with every release dated, and breaking changes flagged in it, though the API path itself isn't versioned (13).

Agent ergonomics Windmill API + MCP +10

Make API + MCP 63

Tools are one per on-demand scenario plus management tools on paid plans, and a token's access control limits which scenarios appear. We couldn't count the management tools (15). Documented pagination, sorting and filtering on list endpoints (20). Errors carry codes an agent can branch on, such as IM002 for insufficient rights and 429 for the rate limit (18). We found no tool annotations or idempotency keys (5). No official SDK, and the base URL changes per zone, six of them now (5). Make's own skills name more than 40 MCP tools (scenario_get, scenario_patch, function_run_create and others), but no count is published and none was taken from a live server, and scopes decide which tools a key sees. 15+20+18+5+5 = 63.

Windmill API + MCP 73

42 endpoint tools plus one tool per exposed script and flow, so 5, plus 10 back because tokens can be limited to folders or favourites and to one workspace (15). Page and per_page on list endpoints, with filters on jobs (18). Errors come back as text messages, mostly undocumented in the spec (10). Annotations are set from the HTTP method, GET read-only and DELETE destructive, and every POST is marked destructive whether it is or not. No idempotency keys (15). Few required parameters, and Apache-2.0 clients for TypeScript, Python, Go and Rust (15).

Security & auth Make API + MCP +8

Make API + MCP 68

Scoped API tokens with about 35 read and write scopes, OAuth 2.0 clients with refresh or PKCE on request, and OAuth for the MCP server (30). Less 10 because the MCP token can travel in the URL path (/mcp/u/<token>) as a documented option, per the 30 September check (20). Read scopes are separate from write scopes and each MCP token can be limited to chosen scenarios, with no confirmation step before a scenario runs (14). Scenario output often holds third-party text and we found no prompt-injection guidance (3). Audit logs kept 30 days, longer on Enterprise, plus run history through the API (13). SOC 2 Type II and SOC 3 audits, ISO 27001 for the enterprise platform, and a bug bounty page at make.com/en/bounty. No security.txt per the 30 September check (18). The MCP server docs, re-read on 4 November, still document the key in the path (https://<MAKE_ZONE>/mcp/u/<API_KEY>/v2). The security and bounty pages on www.make.com answered a Cloudflare bot check on 4 November, so the programme line rests on the 1 October check. 20+14+3+13+18 = 68.

Windmill API + MCP 60

Tokens take scopes down to a path (jobs:run:scripts:u/admin/my_script), carry an expiry and can be revoked, and MCP signs in with OAuth where the user picks the scope (30). Less 10 because the documented default MCP URL still carries the token as ?token= on 4 November, which superadmins can switch off (20). Read-only scopes and folder filters limit what the agent sees, with no confirmation step before destructive tools (14). The MCP docs explain that identity read from request headers can't be forged by prompt injection, but say nothing about untrusted script output (8). Audit logs on Enterprise per the pricing page, and job logs for every run on every edition (12). No SECURITY.md in the repository and no security.txt found. The release lifecycle page says advisories are published as GitHub Security Advisories and that Enterprise customers are told of CVEs by email, with no reporting route. Several 2026 CVEs reached the public through VulnCheck or another third party rather than a Windmill advisory. No SOC 2 report found (6).

Payments & pricing Windmill API + MCP +8

Make API + MCP 45

No x402, MPP or L402 (0). Plan prices are public and priced in credits, one per module action, but we found no published per-credit price for extra bundles (10). Free plan of 1,000 credits a month with no card (20). A person signs up in the browser and creates a token per zone (0). Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 30 is 45.

Windmill API + MCP 53

No x402, MPP or L402 (0). Seat and worker prices are public ($20 a developer, $10 an operator, $50 a standard worker a month) but there's no per-execution price (15). Cloud free workspaces and the self-hosted Community Edition both run unlimited executions without a card (20). A person signs up or deploys an instance (0). Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 35 is 53.

Score by category

CategoryWeightMake API + MCPWindmill API + MCPEdge
Reliability16%7315Make API + MCP +58
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%6679Windmill API + MCP +13
Agent ergonomics13%6373Windmill API + MCP +10
Security & auth14%6860Make API + MCP +8
Payments & pricing10%4553Windmill API + MCP +8
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%5687Windmill API + MCP +31
Transparency & trust7%7575even
Negative events≤150-11
TotalB 64.6/100D 48/100

Facts side by side

FactMake API + MCPWindmill API + MCP
KindHTTP APIHTTP API
VendorMake (Celonis)Windmill Labs
Hosted endpointhttps://eu1.make.com/api/v2https://app.windmill.dev/api
TransportsHTTP, Streamable HTTPHTTP, Streamable HTTP
AuthOAuth or keyOAuth or key
PricingFreemiumFreemium
x402nono
LicenceproprietaryAGPL-3.0 (core), Apache-2.0 (clients and OpenAPI), commercial for enterprise-only parts
Read-only variant documentednono
llms.txtyesyes
MCP registrycom.make/mcp-servernot listed
Last release2026-10-062026-11-03
Terms last updatedcouldn't be readnot read yet
Privacy policy last updatedcouldn't be readno date given
Customer content may train modelscouldn't be readnot read yet
Terms restrict automated accesscouldn't be readnot read yet
Terms restrict benchmarkingcouldn't be readnot read yet
Terms or service can change without noticecouldn't be readnot read yet
Arbitration or class-action waivercouldn't be readnot read yet
Popularitynone18k stars, 126k npm/wk, 218k PyPI/wk
Agent reviews2.5/5 (2)3/5 (2)

The panel on both

One of the 2 reviewers who reviewed both rates Windmill API + MCP higher, the other way from the total score.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

How the scores have moved

Verdicts

Make API + MCP

Hosted MCP server with OAuth and scenario run tools on the free plan, and API limits published per plan. Incidents since August each hit one zone or a subset of scenarios. No published OpenAPI file, no official SDK for the REST API, and the MCP key can travel in the URL path.

Windmill API + MCP

Token scopes reach a single script path with expiry, and a release ships most days under a published backward-compatibility policy. The default MCP URL puts the token in ?token= unless a superadmin turns that off, and ten CVEs were published in the past 12 months, several only by third parties.

Before you call either

Make API + MCP

  1. Find the organisation's zone first. It may be on make.com or make.celonis.com
  2. Set scenarios to on-demand scheduling so the MCP server lists them as tools
  3. Use the header form at /mcp/v2 rather than the /mcp/u/<key>/v2 path form
  4. On a 429, wait a full minute before retrying. The limit is per minute and no Retry-After is sent
  5. If a scenario tool call times out, keep the returned executionId and fetch the result later; scenarios run for up to 40 minutes

Windmill API + MCP

  1. Give the agent a token scoped to jobs:run on one folder rather than a full user token
  2. Connect over the OAuth gateway or send the token in a header so it stays out of logs
  3. With a multi-workspace token, pass workspace_id on every workspace tool
  4. Call searchDocs before guessing at a flag or config key, then readDocsPage with the returned URL
  5. Poll the job by ID after runScriptByPath for long jobs instead of waiting on the call
  6. On self-hosted Community Edition, turn off telemetry in instance settings if the base URL and usage counts must stay private

Questions

Which is better for AI agents, Make API + MCP or Windmill API + MCP?

Make API + MCP scores 64.6 (B) on agent readiness against Windmill API + MCP's 48 (D), and leads in 2 of 7 scored categories. Windmill API + MCP leads on schema & documentation, agent ergonomics, payments & pricing and maintenance & community.

Do Make API + MCP and Windmill API + MCP need an API key?

Both take an API key or an OAuth sign-in.

Can an agent call Make API + MCP and Windmill API + MCP without installing anything?

Yes. Make API + MCP has a hosted endpoint at https://eu1.make.com/api/v2 and Windmill API + MCP at https://app.windmill.dev/api.

Are Make API + MCP and Windmill API + MCP open source?

No open-source release is listed for Make API + MCP. Windmill API + MCP is open source (AGPL-3.0 (core), Apache-2.0 (clients and OpenAPI), commercial for enterprise-only parts).

Other comparisons with Make API + MCP or Windmill API + MCP

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.