Verdict: 6.7/10 (provisional)
Gumloop is a powerful way to hand repetitive admin to AI agents without writing code. But it costs more than the others here, it's aimed more at teams, and there's a learning curve before it pays off.
| Criterion | Score |
|---|---|
| Ease for beginners | 6 |
| Value for money | 6 |
| Usefulness for building a business | 8 |
| Reliability | 7 (provisional, not yet tested) |
| Privacy and data handling | 6 |
| Support | 7 |
| Overall (average) | 6.7 |
- Who it's for: small-business owners who already know which repetitive jobs they want to hand over, such as sorting leads, drafting replies or updating spreadsheets.
- Who should skip it: absolute beginners who haven't yet worked out their processes, and anyone on a tight budget.
- Best thing: AI agents that connect to the tools you already use, with no code.
- Biggest worry: there's no free plan, only a 14-day trial that needs a card and turns into a paid plan, and self-serve customers' content may be used for AI training.
- Bottom line: Use the 14-day trial to automate one specific task, and cancel before it ends unless it's saving you real hours.
What it does, in plain English
Gumloop lets you build AI "agents" and workflows that work across your apps. Gumloop's own example use cases include lead generation, lead qualification, content creation, meeting prep, Shopify stores and support. You describe or assemble what should happen, and it runs on its own. It raised a $17M Series A in January 2025.
Pricing
Converted at £0.7492 per US$1 (ECB reference rates, 26 Oct 2026). Gumloop charges in USD, so your bank may add conversion fees, and VAT may be added. Checked on Gumloop's pricing page and documentation on 27 Oct 2026.
| Plan | Price | What you get |
|---|---|---|
| Pro | from ~£27.72/month ($37), with a 14-day free trial | Unlimited agents, 20,000 credits a month, 35+ AI models, unlimited seats, email support, Gumloop University |
| Enterprise | Custom | Company data features, admin and security controls, dedicated Slack support |
There's no free plan. Gumloop's documentation says every new account starts with a 14-day free trial of Pro. The trial needs a card, is a one-off, and rolls into a paid Pro subscription when it ends unless you cancel first.
How the credits and fees work. One credit is worth $0.005. Each agent run is charged for the AI model's tokens, any paid tool calls (at least 1 credit each) and compute time (5 credits per minute of active work). On top of that, Gumloop adds an 8% "orchestration fee" on the full cost of the run. Its pricing FAQ says the fee rises to 16% if you bring your own AI model API keys. Unused credits don't roll over on Pro. If you run out mid-month your agents stop, unless you switch on pay-as-you-go overage, which is billed at $0.005 a credit and capped at $5,000 per billing period by default (you can set a lower cap).
What it really costs a beginner: at least £27.72 a month after the trial, plus any overage you switch on.
Pros and cons
Pros
- Unlimited seats on Pro, so you can bring in a VA or a partner without extra cost.
- Lots of ready-made use cases to copy.
- Lets you choose from 35+ AI models, or bring your own AI keys (though the orchestration fee doubles to 16% if you do).
- Independent SOC 2 Type II audit.
- Free learning material (Gumloop University, workshops).
Cons
- More expensive to start with than the other tools in this batch, and there's no free plan.
- Building good automations takes thought, so the first week is a learning curve.
- Self-serve content may be used for AI training.
- The 8% orchestration fee (16% with your own keys) is added to every run, which makes costs harder to predict.
Ease for beginners
It's no-code, but you still need to think like a process designer: what triggers the task, what the AI should do and where the result goes. That's a step up from "type a wish, get an app". The good news is that the templates and Gumloop University lower the barrier, and live workshops are advertised. Start with one small, boring task.
Privacy and data handling
- AI training: Gumloop's privacy policy says it "may use Self-Serve Tier Content to develop, train, and improve our AI models". Enterprise content isn't used unless you agree in writing. We found no opt-out for self-serve users (unverified).
- Model providers: Gumloop says it has agreements with OpenAI and Anthropic that they won't train on data sent through Gumloop's API.
- Google data: it says data from Google Workspace APIs isn't used to train general AI models.
- Security: a SOC 2 Type II audit, with the report available on request via its trust centre. Its policy also says it follows the EU–US Data Privacy Framework and its UK extension.
How to start in 15 minutes
- (2 min) Start the 14-day trial at gumloop.com. You'll need a card, and the trial turns into a paid plan unless you cancel, so set a calendar reminder for day 12.
- (3 min) Pick one task you do every week, such as "summarise new enquiries from my inbox into a Google Sheet".
- (8 min) Find the closest template under Use Cases, connect only the accounts it needs, and run it once on test data.
- (2 min) Check the output carefully before letting it run on its own.
Sources
- Gumloop pricing: https://www.gumloop.com/pricing
- Gumloop documentation, credits and plans: https://docs.gumloop.com/core-concepts/credits
- Gumloop documentation, subscription management: https://docs.gumloop.com/core-concepts/subscription_management
- Gumloop privacy policy: https://www.gumloop.com/privacy-policy
- Gumloop trust centre: https://trust.gumloop.com/
- Gumloop $17M Series A: https://www.gumloop.com/blog/gumloops-17m-series-a
- The Logic on the Series A: https://thelogic.co/briefing/vancouver-founded-automation-startup-gumloop-raises-us17m-series-a/
- Gumloop Creators (affiliate) programme: https://www.gumloop.com/partners/apply-to-be-a-creator
- ECB euro reference rates: https://www.ecb.europa.eu/stats/policy_and_exchange_rates/euro_reference_exchange_rates/html/index.en.html