Make
Make is a no-code automation tool where you build workflows, called scenarios, on a visual canvas: modules for 3,500+ apps joined by routers, filters and loops, with AI agents that run inside the same scenario.
It suits founders whose workflows branch, loop or run thousands of times a month, and who want to see the whole flow at once.
It is free for 1,000 credits a month; Core starts at $16 a month for 10,000 credits, or $11 a month billed yearly, as checked on 30 October 2026.
Reviewed by the IndieHacked team
- What it does
- Build automations on a visual canvas that links 3,500+ apps, with routers, loops and AI agents inside the same workflow.
- Our take
- Founders whose workflows branch, loop or run thousands of times a month. Lay the whole flow out on a canvas, put AI agents in the same scenario, and pay from $16 a month for 10,000 credits, with no limit on users.
- Pricing
- FreemiumLow priceFree: 1,000 credits a month, 2 active scenarios and runs at most every 15 minutes. For 10,000 credits a month, Core is $16, Pro $28 and Teams $51 billed monthly, or $11, $20 and $36 a month billed yearly; a slider takes Pro and Teams to 8 million credits, at $6,227 and $10,652 a month, as checked on 30 October 2026. Most actions use 1 credit, and a trigger uses 1 each time it checks, even when nothing is new; AI steps on Make's own AI provider and replies from Maia use credits by tokens. Unused credits expire at the end of the term, and Core's reset monthly even on yearly billing. When credits run out, scenarios stop until you buy more, at 25% above your plan's rate per credit, or turn on auto-purchase. A cancelled plan runs to the end of the billing period, then drops to Free.
The IndieHacked breakdown
What it does
Make builds automations as scenarios: modules for 3,500+ apps placed on a canvas and joined by routers, filters, iterators and aggregators, so one trigger can split into branches and work through a list row by row. The HTTP module and custom apps reach any service with an API, and on paid plans the Make Code app runs JavaScript or Python.
AI sits in the same scenarios. Make AI Agents, in beta on every plan, take a goal, choose among the modules, scenarios and MCP tools you give them and show their reasoning; Maia, in public beta, builds, edits and fixes scenarios from a prompt. Make's MCP server lets Claude or ChatGPT run your scenarios as tools, and Make Grid, from Core up, maps how scenarios, agents and connections depend on each other.
Who it suits
Founders whose automations go past one trigger and one action: orders routed to different tools by country, a list processed row by row, an agent that sorts support emails before the scenario files them. The canvas shows every module and route at once, and Make's own Zapier comparison says most scenarios take unlimited modules and routes, where a Zap stops at 100 steps.
It also fits a small budget at volume. Core's 10,000 credits a month cost $16, and every plan includes unlimited users, so a cofounder or freelancer builds in the same organization at no extra charge. Through the MCP server and Make Skills, in beta, Claude Code or ChatGPT can build and run scenarios for you.
Not the right fit for
Founders new to automation who need one app to tell another about an event, or who rely on less common tools. Make's own comparison says Zapier "works well for simple, linear workflows," and Zapier, reviewed here, lists 9,000+ apps to Make's 3,500+.
Developers who want to host the automation tool themselves or keep it in their own code. Make runs in its own cloud, on AWS in the EU or North America, on every plan; its comparison with n8n names that tool for organizations with plenty of technical staff, as it is self-hosted on most plans and has a free Community Edition.
Also on the shortlist
- ZapierAutomation9,000+ apps in a guided step-by-step editor, with agents, chatbots and an MCP server as products of their own, from $29.99 a month for 750 tasks.Free for 100 tasks a month, paid plans from $29.99 a month, as checked on 30 October 2026
Report an error. Spotted a factual or language mistake? .