Gumloop
AutomationPaidVisit Site
A visual AI automation platform where the nodes are AI steps rather than API calls — closer to an agent builder than to Zapier.
Overview
Gumloop occupies the space between Zapier (great at moving data, weak at judgment) and a code-first agent framework (powerful, requires engineers). You build on a canvas, but the primitives are AI operations — extract, summarize, classify, decide, browse — chained across your tools. That makes it a strong fit for the marketing and ops work that is 80% judgment and 20% plumbing: ad campaign QA, lead research and scoring, content pipelines, CRM hygiene. Shopify, Instacart, and Gusto are named users. Two pricing details matter more than the headline $37/mo: the plan is credit-metered at 20,000 credits, and there is an 8% orchestration fee on top. Model the credit burn on a real workflow during the 14-day trial before you commit, because credit-metered AI platforms are where surprise bills live.
Pros & Cons
Pros
- AI-native nodes, not just data plumbing
- Unlimited agents and seats even on the Pro plan
- 35+ models selectable per step
- Enterprise tier hosts and proxies MCP servers
Cons
- Credit metering makes cost hard to predict — test before committing
- 8% orchestration fee on top of the plan price is easy to miss
- Visual canvases get hard to maintain past a certain complexity
- Smaller connector library than Zapier or Make
Workflows that use Gumloop
Get a new AI workflow each week — many feature Gumloop and other tools in this category.