Retool's Team plan costs $10 a month per builder and $5 per internal user on annual billing, according to its pricing page in October 2026. For two builders and ten users, that is $70 a month, about what one hour of a median US software developer's salary costs. That comparison is why most internal tool development should start in a builder, and why the interesting question is when it should not.
This guide compares three routes (a builder such as Retool or Appsmith, Webflow with Wized and Xano, or custom code), their cost at three team sizes and the signals that move a tool between them. Prices come from vendor pages read on 2026-10-28; salary data from the US Bureau of Labor Statistics.
The three routes, and what each one is underneath
Internal tool development covers the admin panels, approval queues, dashboards and data-entry screens a company builds for its own staff. Retool's 2022 survey put the share of developer time spent on them at 33%, and our internal tools statistics page explains why that number leans high. The three routes differ in where the code lives and who maintains it.
An internal tool builder. Retool, Appsmith, ToolJet and Budibase give you a canvas of tables, forms and charts that connect straight to your databases and APIs. You write queries and small scripts; the vendor ships components, hosting and user management.
Webflow plus Wized and Xano. Webflow draws the interface, Wized wires its elements to data and login state in the browser, and Xano provides the database, API and business logic. It suits teams who already run a Webflow site and want a branded tool.
Custom code. A React or Next.js app on a backend such as Supabase or your own API. The data model, permissions, tests and maintenance are all yours.
The question that decides it: who uses the tool
Start with the user list, not features: it decides the pricing model, the security model and how polished the interface must be.
| Signal | Builder (Retool, Appsmith) | Webflow, Wized and Xano | Custom code |
|---|---|---|---|
| Users are staff only, under about 50 | Best fit | Works, more set-up | Usually overkill |
| Users include customers or partners | Possible on paid tiers, billed per external user | Good fit, no per-user fee | Good fit |
| Interface must match the brand | Limited theming | Full design control in Webflow | Full control |
| Mostly tables, forms and approvals over existing data | Best fit | Slower to build | Slower to build |
| Own data model with complex rules | Possible, logic spreads across queries | Logic lives in Xano | Best fit |
| SSO, audit and Git needed | Enterprise tiers on Retool; Business and above on Appsmith for audit | Build or buy per piece | Build it once |
Source: Retool, Appsmith, Wized and Xano pricing pages, read 2026-10-28; our assessment.
What each route costs at three team sizes
Builders charge per seat, Wized per domain with a page-view cap, Xano per plan, and custom code in developer hours.
| Plan | Builder seat | Internal user | Notable limits |
|---|---|---|---|
| Retool Free | $0 | $0 | Up to 5 users, 500 workflow runs, production only |
| Retool Team | $10 yearly, $12 monthly | $5 yearly, $7 monthly | Staging and production, no external users |
| Retool Business | $50 yearly, $65 monthly | $15 yearly, $18 monthly | External users, audit logs, unlimited environments |
| Appsmith Free | $0 | $0 | Up to 5 cloud users, Git with 3 repos, Google SSO |
| Appsmith Business | $15, no extra charge for builders | $15 | Up to 99 users, audit logs, custom roles |
| Appsmith Enterprise | From $2,500 a month | 100 users included | SAML and OIDC SSO, SCIM |
Source: retool.com/pricing and appsmith.com/pricing, read 2026-10-28. Vendor-sourced figures.
Applied to three teams, on annual billing and without overages or add-ons:
| Team | Retool Team | Retool Business | Appsmith Business | Wized Small and Xano Essential |
|---|---|---|---|---|
| 2 builders, 10 users | $70 | $250 | $180 | About $114, plus Webflow plan |
| 3 builders, 40 users | $230 | $750 | $645 | About $114, plus Webflow plan |
| 5 builders, 120 users | $650 | $2,050 | Enterprise, from $2,500 | About $114, plus Webflow plan |
Source: retool.com/pricing, appsmith.com/pricing, wized.com/pricing and xano.com/pricing, read 2026-10-28. Wized Small is $29 to $39 a month depending on billing and caps a domain at 100,000 page views and 100 elements; Xano Essential is $85 yearly or $99 monthly.
The Wized and Xano column is flat because neither charges per user, but someone has to build login, roles and every screen, which a builder gives you on day one. Our Wized vs custom code comparison covers where that build time goes. External users need Retool Business, where the first 50 are free and the listed rate then falls from $8 to $4 a user a month as the count rises.
Custom code in hours. The BLS puts the median software developer wage at $135,980 for May 2025. Over 2,080 working hours that is about $65 an hour before tax, benefits and office costs. Supabase Pro starts at $25 a month with 100,000 monthly active users, so hosting is small; build and upkeep are the cost.
When Retool or Appsmith is enough
A builder is the right default when staff read and write data that already exists: refund approvals, a support order lookup, a moderation queue, a stock adjustment screen.
Speed wins here. A developer who knows the data can have a working screen the same day, with users and roles handled by the platform.
The governance gaps are specific. On Retool, SSO, SCIM and Git source control are Enterprise features, and self-hosting is Enterprise-only, deployed on Kubernetes with Helm for production. On Appsmith, audit logs start at Business, and SAML or OIDC sign-in needs Enterprise. If your security team requires company sign-in for every tool, price the Enterprise tier from the start, not after the pilot.
The open-source option. Appsmith's Community edition is released under the Apache-2.0 licence, has 41,100 GitHub stars and installs with Docker or Kubernetes. It removes the per-seat bill but adds a server someone must patch and back up.
The signal that a builder has stopped being enough is usually not a missing feature. It is logic: business rules scattered across dozens of query panels that only one person understands, with no tests and no review.
When Wized and Xano win
The Webflow route earns its place when the tool faces outsiders or must match the brand: a partner portal, a client onboarding dashboard, a supplier order form.
No per-user bill. At 400 partner logins, Retool's external user bands still apply on Business. Wized prices by domain and page views, and Xano prices by plan, with no platform-enforced API rate limit on Essential and above.
Design control. Webflow gives a designer the whole canvas; a builder gives you themed components, which staff accept and customers notice.
The trade-offs. Logic lives in Xano, so the backend is visual too, and Xano's free plan allows only 10 API requests every 20 seconds, which is for building, not for users. Role-based access inside Xano's own team workspace starts at Pro ($224 a month yearly). HIPAA is a paid add-on on Pro or included on Custom. Self-hosting needs the Custom plan. Our Xano vs Supabase comparison sets out those limits in detail.
We would recommend it mainly to teams already running their marketing site in Webflow.
When to write the internal tool in code
Custom code is the slowest start and the cheapest ceiling. Six signals tell you the tool has reached it.
- The data model is the product. If the tool defines new records, states and relations rather than editing existing ones, it needs migrations and tests.
- Permissions depend on the data. "Account managers see only their own clients" is row-level security, which belongs in the database, not in a front-end condition.
- Customers will use it at scale. Per-user fees and builder theming both hurt at hundreds of outside users.
- The builder bill rivals a developer. At $2,050 a month, Retool Business for 125 people costs $24,600 a year, about 375 hours of median developer salary. That buys the first version of most admin tools, though not the upkeep.
- Review and rollback matter. If changes must go through pull requests and staging, code gives you that natively; Retool reserves Git for Enterprise.
- The tool may become a product. Internal tools that work tend to get shown to customers. Code moves with you; builder apps mostly do not.
Two or three together usually make the case. Retool's own 2026 Build vs. Buy report, a survey of 817 of its customers and builders, found 35% had already replaced at least one SaaS tool with a custom build, and 78% planned to build more. It also found engineering bandwidth was the top blocker at 42%, which is the honest cost of this route.
For the stack itself, our MVP tech stack post explains the Next.js and Supabase pairing we would start from, and why.
Permissions, audit and upkeep: what every route needs
OWASP ranks broken access control first in its 2021 Top 10: 94% of applications tested showed some form of it. A login does not exempt internal tools. In Retool's 2022 survey, over a quarter of respondents said their internal tools had no access controls at all.
Whatever the route, scope these five things before the first screen.
Roles. Who can view, who can edit, who can approve. Keep it to three or four roles.
Where the rule lives. A hidden button is not a permission. The check has to run on the server or in the database.
Audit. Who changed which record and when. Builders include this on paid tiers; in code it is a table and a trigger.
An owner. One named person who answers for the tool when its builder leaves.
A retirement date for the spreadsheet. If the old sheet stays editable, people will keep using it.
What this means for choosing your internal tool route
Start with the user list and the budget line, not the feature list. If the users are staff and the data already exists, open a builder and ship a screen this week. If outsiders will log in or the brand matters, and your marketing site already runs in Webflow, price Wized and Xano. If the tool defines its own data, needs row-level permissions or is likely to become a product, write it in code from the start rather than migrating later.
We build applications in Next.js and marketing sites in Webflow, so read our view with that bias in mind. Our web apps service page explains how we scope a custom internal tool and quote it per project.
