Make review: visual automations with AI agents you can watch think
The most transparent automation canvas in this guide, with agents whose every decision is visible. Governance features and readable pricing are the gaps.
Contents
| Type | Visual automation platform |
|---|---|
| Pricing | Published |
| Website | make.com |
Make builds automations as visual scenarios: modules chained across more than 3,000 apps. Its AI agents live on the same canvas, so an agent is one more visible part of a scenario rather than a separate product1.
Best for: Teams that want to see and control every step of an automation, including AI agents whose reasoning is shown step by step.
Not the right fit for: Buyers who need audit logs, SSO or custom roles without an Enterprise plan.
What Make's AI agents do
Make AI Agents are reusable across scenarios, and a reasoning panel shows each decision step by step. Builders can add manual approvals or stop an agent at specific points, and AI agents are available on all plans1. Agents can use Make's own AI provider or, on paid plans, the company's own provider connection2.
How credits work
Make bills in credits; by default one operation costs one credit. With Make's AI provider, an agent costs one credit per operation plus credits based on the model tokens used; with the company's own provider connection, it costs one credit per operation and the provider bills the tokens23. The free plan includes 1,000 credits a month3. Since November 2025, Core allows up to 300,000 credits a month and Pro up to 8 million4. Extra credits cost 25% more than plan credits and expire4. The prices of paid plans were not readable on the pricing page when we checked5.
Governance
Audit logs, kept for 12 months, are available on the Enterprise plan only6, as are custom roles7. An on-premise agent that lets scenarios reach systems on a local network also requires Enterprise8.
Published evidence
- TripleTen. Make reports that TripleTen cut ad setup time by up to seven times and moved creative labelling for more than 1,000 assets a year from weeks to seconds, using an AI agent among its scenarios. Vendor-reported.9
Strengths and limits
Strengths
- A reasoning panel shows each decision an agent makes, which makes errors easier to find.
- AI agents, including manual approvals and stop points, are available on all plans.
- Paid plans can bring their own AI provider, so model costs are billed by the provider rather than in Make credits.
Limits
- Audit logs, SSO and custom roles are Enterprise-only.
- Plan prices could not be read on the pricing page when we checked, so costs must be confirmed directly.
- Extra credits cost 25% more than plan credits and expire.
- No self-hosting; an on-premise agent for reaching local systems requires Enterprise.
Choose another option if: You need governance features on a mid-market plan, or a system that holds the records itself. Zapier or OpenLazarus fit those needs.
Pricing
Credit model published; free plan 1,000 credits a month. Paid plan prices were not readable on the page we checked.
Questions to ask Make
- What does each paid plan cost for our expected monthly credits?
- Should our agents use Make's AI provider or our own, and how does that change the bill?
- Which steps should stop for manual approval?
- Do we need audit logs or custom roles, and therefore Enterprise?
See how Make compares with the other tools in the main ranking.
Sources
- Make AI Agents: build transparent AI agents across 3,000+ apps. Make. Accessed 8 Nov 2026.
- Credit usage for AI agents. Make Help Center, 10 Jun 2026. Accessed 8 Nov 2026.
- Credits. Make Help Center. Accessed 8 Nov 2026.
- Adjustments to plans and pricing. Make Help Center, 6 Nov 2025. Accessed 8 Nov 2026.
- Pricing and subscription packages. Make. Accessed 8 Nov 2026.
- Audit logs. Make Help Center. Accessed 8 Nov 2026.
- Custom roles. Make Help Center, 25 May 2026. Accessed 8 Nov 2026.
- On-premise agent. Make Help Center. Accessed 8 Nov 2026.
- TripleTen cuts ad setup time 7x with Make and Make AI Agents. Make, 18 May 2026. Accessed 8 Nov 2026.