Gumloop
A node-based canvas where each step is an AI action — research, classify, extract, write — chained together into one automated workflow.
What it costs
| Tier | Price | For | Includes |
|---|---|---|---|
| Free | $0 | Trying the canvas and building a first working flow. | A 14-day free trial of the paid tier's full feature set before dropping to free-tier limits. |
| Pro | $37/month | One person or a small team running AI workflows regularly. | 20,000 credits a month (7,400 base plus 12,600 bonus), covering model tokens, compute and paid API tool calls at each provider's list price plus an 8% orchestration markup. |
Read from Gumloop's own pricing page on 25 Sep 2026. Vendors change prices without notice.
Where it breaks
Higher tiers (Team, Enterprise) exist based on product structure but their prices were not found in the page's structured pricing data and are not stated here. Credits are consumed by underlying model and API costs at each provider's own rate plus an 8% markup, so the real monthly cost depends entirely on which nodes a workflow uses — the $37 buys an allowance, not a fixed feature set.
Entirely browser-based, no install. Connects to 30+ external services (Apollo, Exa, Parallel and others) as workflow steps, each billed through Gumloop's unified credit system rather than separate accounts.
First entry in this reference. There is no previous check to compare it against.
Gumloop does the same job as Make — a workflow that runs on a schedule or trigger, not something you sit and operate — but built AI-native from the start rather than having AI steps bolted onto an older automation product. Every node on the canvas can be a model call: summarise this, classify that, extract structured data from this messy input, and pass the result to the next step.
The credit system is the thing to understand before anything else. Gumloop does not charge a flat per-workflow fee; it passes through the actual cost of whatever the workflow does — a token from an AI provider, a paid API call to a research tool, a unit of compute — at that provider’s list price, plus an 8% fee for handling the billing and orchestration. The $37/month Pro plan buys 20,000 of those credits, not 20,000 workflow runs; a workflow doing expensive research-API calls burns through the allowance faster than one doing simple text classification.
Who this suits
If your automation needs are mostly “move data from A to B on a trigger,” Make or n8n do that more cheaply because they are not billing for model inference on every step. Gumloop earns its place when the workflow itself needs judgment in the middle — read this inbound lead, decide if it’s qualified, draft a tailored reply — the kind of step a rules-based automation tool cannot do at all.
The honest limit
Credit consumption is genuinely hard to predict before you have run a workflow for real, the same warning that applies to Make’s credit model. Build the workflow, run it against real data for a week, then judge the monthly cost — not from the sticker price alone.
Get Gumloop
Buying the tool is not the hard part.
The ebook is about what to hand over, in what order, and how you find out when it goes wrong. One payment, no subscription.