"Self-serve" is doing a lot of work
Every PSD2 account-information provider markets some version of "get started in minutes." What varies — enormously — is what you can actually do before a human from sales gets involved. There are really three gates:
- Signup — can you create an account at all, alone, with an email address?
- Sandbox — can you call working APIs before signing anything?
- Production — can real users at real banks connect, and under what terms?
A provider can be wide open at gates one and two and still be effectively closed at gate three. That's the trap, and it's why "self-serve PSD2" comparisons deserve more than a feature checkbox.
The field as of August 2026
(From actually attempting the signups, not from reading landing pages.)
Enable Banking is the closest thing to genuine self-serve among the challengers. Email magic-link signup, account auto-created, a sandbox with no contract — and, notably, restricted production access with no contract either, where you can connect your own accounts. Their sandbox is genuinely good: real bank coverage, honest documentation, no credit card. Credit where due.
But the moment your app serves other people — public, customer-facing production — you need a signed contract, KYB (know-your-business) verification, and volume-based billing priced by their sales team via info@enablebanking.com. Perfectly reasonable for a business with a finance function. Not the "self-serve" an indie builder was hoping for.
GoCardless Bank Account Data (formerly Nordigen) — closed to new onboarding since 2025, free tier sunset. The platform that built a thousand side projects no longer takes new independent developers.
TrueLayer — free tier gone, /pricing 404s, and a clear retreat upmarket to enterprise. The live-upgrade path involves sending your client_id to sales.
Tink (Visa) — "Contact Sales" on effectively every page.
Yapily — "Talk to an expert." Same wall.
None of this is misconduct. These are businesses that did the math and stopped subsidizing hobbyists. But if you ARE the hobbyist, the practical reading is: the 2021-era free-open-banking window has closed at the big platforms.
The table
| Provider | Signup | Sandbox | Prod (own accounts) | Prod (customers) |
|---|---|---|---|---|
| Enable Banking | Self-serve, magic link | Free, no contract | Restricted, no contract | Contract + KYB + sales-priced volume |
| GoCardless Bank Account Data | Closed to new users | — | — | — |
| TrueLayer | Sales-assisted | Limited | Via sales (send client_id) | Enterprise |
| Tink | "Contact Sales" | Gated | Sales | Sales |
| Yapily | "Talk to an expert" | Gated | Sales | Sales |
| open-banking.io | Self-serve | Yes | Yes — your own bank keys | Yes, from day one (~€3/mo) |
How to check the gate yourself in ten minutes
Before you commit a weekend of integration work to any provider, run this diligence pass. It's what produced the table above.
- Find the pricing page. Does it exist? A 404 on /pricing tells you the company has stopped caring about small buyers — the funnel starts with a sales call now.
- Try the signup with a throwaway email. Magic link in your inbox within a minute = real self-serve. A "we'll be in touch" form = sales-walled, whatever the landing page says.
- Search the docs for "production" or "go-live". Read what the upgrade requires. If the words "contract", "KYB", or "contact our team" appear, budget weeks, not minutes.
- Check who holds the keys. Do you register with banks yourself and use your own client credentials, or does the provider hold everything? Own keys means you can leave later without re-consenting every user through yet another provider.
- Ask what a certificate costs you. If the answer is "nothing, our license covers it", that's the cert-free model. If you're expected to produce a QWAC, that's €2–10k a year you hadn't budgeted.
Ten minutes of this beats three weekends of integration into a platform you'll have to leave.
The decision framework
The real question isn't "who's best" — it's "which gate model does my project need?"
Restricted production (own accounts only) is enough when:
- You're building a personal tool — your own budgeting dashboard, net-worth tracker, tax-prep helper
- You're evaluating integrations before committing
- The only bank accounts that will ever pass through the app are yours
In that world, Enable Banking's model is genuinely a good deal: free, no contract, your real data. I'd use it without hesitation for a personal project.
You need public production the moment:
- Anyone who isn't you connects a bank
- You charge money for the product
- You hand it to a beta group, a friend, or a client
The instant "any real user" exists, restricted-prod models snap you into contracts, KYB paperwork, and sales cycles measured in weeks. Fine if you're at that stage and want negotiated volume pricing. If you're not, it's friction between you and your first ten users — which is where most side projects die.
Price in the certificate tax, too. Going direct rather than through any aggregator usually means eIDAS QWAC/QSealC certificates: €2–10k a year and months of paperwork before your first API call. Some aggregator models (mine included) are structured so the aggregator's license covers that and you never touch a certificate. Know which model you're signing up for before you spend a weekend on integration.
Disclosure
I build open-banking.io — a certificate-free EU/UK PSD2 account-information API. Self-serve signup; you use your own bank-issued client keys, with the aggregator-style license handled for you; about €3/month; new customers are welcome, with public production access from day one. I'm obviously biased, which is why the table above describes competitors as accurately as I can manage — including Enable Banking's genuinely free tier, which fits some of these use-cases better than mine does.
Match the gate model to the use-case. That's the whole article.
Top comments (0)