Make Review 2026: We Built and Ran a Real Scenario, AI Assistant Included
Quick Verdict
// Hands-on tested
Teknexia AI Tool Evaluation Framework
Scored against our site-wide 5-criteria framework, from our own hands-on test building and running a real scenario with Make's AI assistant, combined with Make's public pricing page for the tiers we didn't test directly. This is the canonical score for Make, the third entry in our AI Automation category.
Overall: 7.2/ 10
Make is a visual automation platform: scenarios built from a trigger and a chain of modules, connecting apps and built-in tools without writing code. We tested it directly, including Maia, Make's AI scenario-building assistant, on a real account. This is the third review in our AI Automation category, alongside n8n and Zapier, and it's labeled Hands-on tested for the parts we verified directly.
- Maia, Make's AI assistant, builds scenarios conversationally — it asked us which Slack destination and which timezone to use before writing anything, rather than guessing (Teknexia Hub testing, August 2026).
- We hit two real limits in our own account (no Slack connection configured, no free Data Store storage), and Maia correctly adapted the plan both times instead of leaving us with a broken scenario.
- The final scenario — a daily schedule trigger into a Tools module — ran successfully on the first execution: Success, in under a second, for 1 credit.
- This review was run through direct browser automation rather than manually captured screenshots, so it doesn't include interface images the way our other reviews do — the test itself and its results are still real and hands-on.
What We Tested, and What Happened
We logged into an active Make account and opened a new scenario. Instead of building it manually, we typed a plain-language request into Maia, Make's AI assistant (in beta): “Every day at 9am, send me a Slack message saying 'Daily check-in reminder'.” Maia didn't just generate a guess — it asked us two clarifying questions first: where the reminder should go in Slack (we chose DM) and which timezone “9:00 AM” should use. Our first answer, “local device,” wasn't precise enough, so it followed up asking for a real IANA timezone (we picked Europe/London). It then summarized the exact plan — a daily 9:00 AM Europe/London schedule into a Slack DM — and asked us to confirm before building anything.
Once we confirmed, Maia built the scenario and asked us to connect a Slack account. Our test account had no Slack connection configured, and creating one requires an OAuth authorization we weren't in a position to complete mid-test. Rather than treat that as a dead end, we told Maia directly, and it proposed swapping the action for a Data Store record instead, asking three more clarifying questions (how each record should represent a run, which fields to store, whether to rename the scenario). We agreed, and it built that version — only to run into a second real limit: our account's Data Store storage was full, and creating a new one failed with an explicit “Not enough space in storage” error. We told Maia that too, and it adapted again, this time swapping the action for a built-in Tools “Set variable” step that needs no external account and no storage.
That final version worked. We ran it manually, and Make's own history log recorded the result plainly: Success, triggered manually, completed in under a second, 1 operation, 1 credit used, 67 bytes of data transferred. The variable it set contained exactly what we asked for, composed with a real timestamp: “Daily check-in reminder — 2026-08-15 09:13.”
Our Test Protocol
These notes document what we did during testing — they are separate from the 5-criteria Teknexia AI Tool Evaluation Framework score in the Quick Verdict above, not a replacement or duplicate of it.
| Test | Prompt | Result | Note |
|---|---|---|---|
| AI scenario building (Maia) | “Every day at 9am, send me a Slack message saying 'Daily check-in reminder'” | Asked clarifying questions, confirmed the plan, then built the scenario | Conversational, not one-shot |
| Adapting to a missing Slack connection | Told Maia no Slack connection existed | Proposed and built a Data Store alternative | Handled correctly |
| Adapting to a Data Store storage limit | Data Store creation failed: "Not enough space in storage" | Proposed and built a Tools "Set variable" alternative, no account or storage needed | Handled correctly |
| Scenario execution | Manual run of the final scenario | Success, under 1 second, 1 operation, 1 credit, correct output value | Fast, correct on the first real run |
A note on this review's format
Every other review in this category was captured by the person testing the tool, with their own screenshots. This one is an exception: we ran the entire test ourselves through direct browser automation rather than manual screenshots, so this review doesn't include interface images the way our n8n and Zapier reviews do. The data above — the questions Maia asked, the two real errors we hit, the exact execution result — is all real and drawn directly from that session, not reconstructed from memory or invented.
What we didn't test
We didn't test a real third-party app connection (Slack or otherwise) end to end, since our account had no connection configured and we didn't complete a new OAuth authorization mid-test. We also haven't tested Make's paid tiers, its error-handling and retry tools, webhooks, or its broader module library ourselves. For those, this review leans on Make's own documentation rather than our own use, and we'll expand it as we test more of the platform.
What Make Actually Is
Make (formerly Integromat) is a visual automation platform: scenarios are built on a canvas from a trigger and a chain of modules, connecting apps and built-in tools (schedulers, data stores, HTTP requests, flow control) without writing code (Make, 2026). Maia, its AI scenario-building assistant, is currently in beta and sits inside the same scenario editor, building and modifying scenarios directly on the canvas based on a conversation rather than a single prompt.
Make Pricing and the Free Plan, in Detail
We verified this pricing directly on Make's own /en/pricing page on 15 August 2026. Make switched from counting "operations" to "credits" in 2025 — our own test run used 1 credit for one module execution.
| Plan | Price | Credits/mo | Key limits |
|---|---|---|---|
| Free | $0 | 1,000 | Up to 2 active scenarios |
| Core | $9/mo | 10,000 | Unlimited active scenarios |
| Pro | $16/mo | 10,000 | Advanced scenario controls, priority execution |
| Teams | $29/mo | 10,000 | Team collaboration features |
| Enterprise | Custom | Custom | High-performance execution, dedicated support |
What Users Report
Trustpilot's own site returned an automated bot-detection challenge when we tried to verify Make's exact rating directly, the same block we've run into reviewing other tools in this category. Third-party sources report roughly 164 contributors with mixed sentiment; we don't have a confirmed aggregate star rating to report, so we're not inventing one.
What we can report, cross-referenced across multiple independent sources, are recurring patterns:
- Recurring positives: reviewers describe Make as powerful and flexible once learned, with genuine time savings for recurring workflows — consistent with what our own test found once we got past the two account limits.
- Recurring complaints: a real learning curve around field mapping and branching logic for new users, and inconsistent response times from customer support in some reports — we didn't test support ourselves.
Make vs Zapier vs n8n
All three tools in this category are reviewed here as Hands-on tested, which makes this a genuine three-way comparison rather than one tested tool against marketing copy for the others.
| Criteria | Make | Zapier | n8n |
|---|---|---|---|
| Framework score | 7.2 / 10 | 7.4 / 10 | 7.2 / 10 |
| Free plan | 1,000 credits/mo, confirmed hands-on | 100 tasks/mo, 2-step Zaps only | Free self-hosted only |
| Entry paid price | $9/mo (Core) | $19.99/mo (Professional, billed annually) | $20/mo (Cloud) |
| AI scenario builder | Maia (beta) — conversational, asks clarifying questions, confirmed hands-on | Copilot — single-prompt draft in under 30 seconds, confirmed hands-on in our Zapier review | Has an AI Assistant per n8n's own docs; not tested by us |
| AI output quality in our tests | Every proposal was logically correct; adapted twice to real account limits | Picked the right apps in most tests, but field mapping often needed manual correction | Not tested |
| Review method here | Hands-on tested | Hands-on tested | Hands-on tested |
Pros & Cons of Make (2026)
Pros
- Maia asks clarifying questions before building anything, which made it hard for it to guess wrong in our test
- Adapted correctly to two real account limits (missing Slack connection, full Data Store) without producing a broken scenario either time
- Our test scenario ran successfully on the first try: under 1 second, 1 credit
- 1,000 free credits a month is a genuinely usable free tier, confirmed hands-on
- Cheapest entry paid tier ($9/mo) of the three automation tools we've reviewed in this category
Cons
- Maia's conversational approach takes more back-and-forth than Zapier Copilot's single-prompt draft, in our own test of both
- We didn't get to test a real third-party app connection end to end, since our account had no Slack connection set up
- The identical 10,000-credit figure across three different paid tiers on the pricing page is confusing enough to warrant double-checking before subscribing
- We could not confirm Make's Trustpilot rating; Trustpilot blocked our direct access
- This review doesn't include interface screenshots, an exception to how we normally document hands-on tests
Frequently Asked Questions
Does Make have an AI assistant like Zapier Copilot?
Does Make have a free plan?
How fast does a Make scenario run?
How does Make compare to Zapier and n8n?
Our Make Review Verdict
Make's AI assistant is genuinely conversational, and that turned out to matter more than we expected. Maia asked us real clarifying questions before building anything, and when our test account hit two separate real limits — no Slack connection, no Data Store space — it adapted the plan correctly both times instead of leaving us with something broken. The scenario we ended up with ran successfully on the first try: under a second, one credit, the right output. Set against Zapier's faster single-prompt Copilot, Make's approach costs a few extra exchanges but produced consistently correct results in our test, and its free plan and $9/mo entry price are the most generous of the three automation tools we've now reviewed hands-on. The honest gap in this review: we didn't get to test a real third-party app connection ourselves, and we're publishing it without interface screenshots, an exception to how we normally document these tests. This verdict is Hands-on tested for the parts we verified directly, and we'll revisit it with a fuller test, screenshots included, when circumstances allow.