ToolJet vs Superblocks for internal tools: we built the same apps on both

We gave Superblocks and ToolJet the same internal-tool prompts. Superblocks’ AI builder, Clark, makes polished screens and finished both follow-up edits completely, but in our tests its apps kept no shared data, our Teams workspace had no writable database configured, and every run used dollars of AI usage where ToolJet’s used cents: the eight-page app plus two edits came to $53.71 on Superblocks against $0.38 on ToolJet with GPT-6 Luna.

• Cost: the eight-page app and two edits cost $53.71 on Superblocks, $0.38 on ToolJet with GPT-6 Luna and $2.04 with GPT-6.1 Sol: about 140 and 26 times less on ToolJet, valuing each platform’s AI usage at the rates below. The expense tracker cost $9.25 against 3 to 22 cents.
• Data: our Superblocks Teams workspace had no writable database configured. For the expense tracker we chose browser storage when Clark offered it; for the eight-page app Clark chose in-memory storage on its own, so a new vendor and a lease renewal vanished on reload. Every ToolJet build wrote to ToolJet Database and passed every save check.
• Quality and speed: Superblocks’ edits were complete, including filling in costs on past jobs, which ToolJet’s edits skipped, and they were faster: about 5 and 4 minutes against ToolJet’s 7 to 11. Its eight-page build took about 13 minutes; ToolJet’s took 11 to 17.
• Your own data: both connect to Databricks natively and built a live dashboard. ToolJet’s cost $0.11 to $0.87, Superblocks’ $11.70, and Superblocks’ Overview totals were shifted by a time-zone error.
• Plans: ToolJet has a free plan and a $23 Basic plan with 2,000 AI credits per builder. Superblocks has no free plan; Teams sells a shared pool of GAUs, from 100 for $100 a month billed annually, and one app plus two edits used 43 of them.
• Editing: ToolJet’s apps open in a visual builder anyone can change. Clark writes a React codebase (81 files, about 10,500 lines and 55 npm packages for our eight-page app) that you change by prompting again, which spends GAUs, or by editing code.
• Models and agents: ToolJet offers 19 models from five providers and an MCP plugin that builds with your own coding agent without ToolJet credits. Clark has no model picker, and Superblocks’ Builder MCP still builds through Clark.
• Governance: Superblocks includes role-based access on Teams. ToolJet puts SSO and audit logs on its Team plan; Superblocks keeps them for Enterprise.
• Hosting: ToolJet has a free open-source edition you can self-host. Superblocks’ self-managed deployments are Enterprise only.

Both products got the same prompts, word for word apart from one stray word noted at the end, and we judged what each built without touching it by hand. These are the prompts from our Retool, Lovable and Emergent comparisons.

1. Expense tracker (one page, with a design brief): add expenses, filter by category and month, show the month’s total, in a specified look.
2. Property manager (eight pages): dashboard, properties, units, tenants, leases, rent, maintenance and vendors, with seeded data and working forms.
3. Two follow-up edits on the eight-page app: a Reports page with CSV export and costs on past jobs, then a navy theme, money without cents and a new column.
4. Save checks: add an expense; add a vendor and renew a lease; each confirmed after a reload.
5. Your own data: a three-page analytics dashboard on live Databricks tables.

Plans. Superblocks has no free plan, so we used a paid Teams plan and its AI builder, Clark, on default settings; Clark has no model picker, and Superblocks does not say which models it uses. ToolJet ran on GPT-6 Luna, its cheapest model, and GPT-6.1 Sol, a stronger one. The two platforms therefore ran different models, so these results compare the products as a whole and cannot separate the model from the platform.

Dollars. Superblocks meters AI in Governed Agent Units (GAUs): Teams includes 100 a month for $125 billed monthly, so we value a GAU at $1.25, the same as its top-up rate. ToolJet sells 100 credits for $1. Superblocks reports GAUs per Clark turn, so every Superblocks figure is exact. These dollar figures are the value of the AI usage each run consumed, not extra charges per run; the scaling section uses annual prices.

Rules. One run per prompt, no hand edits. When an agent stopped for approval, we approved and counted the stop; time excludes minutes spent waiting on us. When Clark asked how to store data, we chose browser localStorage. This is a same-prompt test, not a benchmark.

Who ran this. ToolJet ran this test. We report where Superblocks did better, and every prompt is at the end. Tested on 5 October 2026.

ToolJet has a free plan and a $23 entry plan with 2,000 AI credits per builder; Superblocks Teams starts at $100 a month for a shared pool of 100 GAUs.

	ToolJet	Superblocks
Free plan	Yes: 100 AI credits a month, 2 builders, 50 end users, 10 apps	None
Entry paid plan	Basic: $23 per builder a month, billed annually, up to 2 builders, 2,000 credits per builder	Teams: a shared organization pool of GAUs, from 100 GAUs for $100 a month billed annually ($125 monthly); no per-builder fee
Value of one AI unit	$0.01 per credit	About $1.25 per GAU
Published apps	Up to 10 on Basic and Pro; unlimited on Team	1 included on Teams; $10 a month for each extra app
Built-in database	ToolJet Database (PostgreSQL)	None writable in our Teams workspace (its demo Postgres is read-only); Superblocks Database runs in your own AWS account on Enterprise hybrid setups
Models for building	19 models from five providers, picked per build	No picker; Superblocks-hosted models
SSO and audit logs	Team plan ($199 per builder)	Enterprise plan
Self-hosting	Free open-source edition; paid plans self-host too	Hybrid or Cloud-Prem deployment on Enterprise

Sources: ToolJet pricing, Superblocks pricing, read on 5 October 2026.

All three trackers saved an expense that survived a reload; ToolJet’s cost 3 to 22 cents, Superblocks’ $9.25, and only ToolJet’s stored it where a whole team can see it.

Build	Time	Cost	Stops	Add-expense check
ToolJet, GPT-6 Luna	4m 18s	$0.03	0	Passed (ToolJet Database)
ToolJet, GPT-6.1 Sol	2m 25s	$0.22	0	Passed (ToolJet Database)
Superblocks	About 4 min of building (18 min in all, about 14 waiting on us)	$9.25 (7.40 GAUs)	3	Passed in the same browser (localStorage)

Clark first planned to store expenses in its “[Demo] Orders” Postgres and asked to create a table there. The demo database is read-only, so the create failed with “permission denied for schema public”, and Clark offered three ways forward: browser localStorage, connecting a writable Postgres of our own, or in-memory state. We chose localStorage. The finished tracker matched the brief’s layout and colours, but its total card shows all-time spending rather than the month’s total the prompt asked for.

All three builds delivered the eight pages; both ToolJet apps kept their saves, while Superblocks’ app lost them on reload, and cost $19.40 against 21 and 79 cents.

Build	Time	Cost	Stops	Pages	Where the data lives	Save checks
ToolJet, GPT-6 Luna	17m 24s	$0.21	0	8 of 8	ToolJet Database	Vendor and renew passed
ToolJet, GPT-6.1 Sol	11m 11s	$0.79	0	8 of 8	ToolJet Database	Vendor and renew passed
Superblocks	About 13 min	$19.40 (15.52 GAUs)	2	8 of 8	In-memory store	Vendor and renew worked until a reload, then both were gone

Where Superblocks put the data. This time Clark did not ask: it wrote the seed data into the app’s code and kept every change in an “in-memory reactive store”, in its own words. A vendor we added appeared in the list, and a lease renewal moved its end date from Sep 30, 2026 to Sep 30, 2027, but after a reload both were back to the seed data. The screens are polished; the app holds no data between visits.

One note on fairness: our first Superblocks attempt at this app is excluded, because our test harness sent the first edit before the build had finished. The run reported here is a clean rerun.

Superblocks made both edits completely and faster; each cost about 200 times ToolJet’s GPT-6 Luna edit.

Edit	Build	Time	Cost	Result
Reports page, chart, table, CSV export, costs on past jobs	ToolJet, GPT-6 Luna	10m 26s	$0.08	Page and export built; past jobs’ costs not filled in
	ToolJet, GPT-6.1 Sol	8m 1s	$0.65	Page and export built; past costs left behind a one-click fill button
	Superblocks	About 5 min	$15.71 (12.57 GAUs)	Page and export built; past jobs’ costs filled in
Navy theme, whole dollars, days-to-lease-end column	ToolJet, GPT-6 Luna	11m 13s	$0.09	All three changes
	ToolJet, GPT-6.1 Sol	7m 25s	$0.61	All three changes
	Superblocks	About 4 min	$18.60 (14.88 GAUs)	All three changes

Each Superblocks edit stopped once for plan approval. Its Reports chart combines all properties, with the per-property numbers in the table, and one lease that ended on Sep 30 still reads “Expiring”. The edits ran on the same in-memory app, so the costs Superblocks filled in also disappear on reload.

Both platforms connect to Databricks natively and built a working live dashboard; ToolJet’s cost $0.11 to $0.87, Superblocks’ $11.70.

We connected the same Databricks workspace to both, using Databricks’ sample Bakehouse sales tables, and asked for a three-page sales analytics dashboard: KPIs with period-over-period change, revenue by day with a 7-day moving average, hour and weekday patterns, a franchise scatter chart, filters on every page and a full design brief, all queried live with no seeded data.

Build	Time	Cost	Stops	Result
ToolJet, GPT-6 Luna	12m 49s	$0.11	0	All 3 pages on live data; each page queries when you click Load
ToolJet, GPT-6.1 Sol	9m 13s	$0.87	0	All 3 pages on live data, queried on open; default date range a day early
Superblocks	About 9.5 min	$11.70 (9.36 GAUs)	1	All 3 pages on live data, queried on open; Overview totals shifted by a time-zone error

Every app read the warehouse correctly on its Franchises and Customers pages. ToolJet’s GPT-6 Luna build showed the true totals, $66,471 of revenue from 3,333 transactions; Superblocks’ Overview showed $65,538 from 3,276, because a time-zone shift moved its date boundaries, and its period-over-period tags read “+6,924%” against a prior period holding a sliver of data. Clark noticed the shift during its own check and called it expected. ToolJet’s GPT-6 Luna build used ToolJet’s load-on-click option for warehouse queries, so opening it spends no Databricks compute until someone asks for the data; its one flaw was a franchise revenue chart drawn unsorted, and GPT-6.1 Sol’s was a default date range a day early.

Setup. Both take the warehouse details and a personal access token; ToolJet’s connection test passed first time.

Each screenshot is the app exactly as the AI left it, with no hand edits, labelled with the model or plan, what the run cost and how long it took.

Superblocks’ screens are polished. What a screenshot cannot show is where the data goes and what each run cost: every ToolJet app here keeps its data where the whole team shares it, for cents. Superblocks’ expense tracker kept data in one browser and its property manager kept data in memory, for dollars per run; both platforms’ Databricks dashboards read the connected warehouse.

Clark writes a React and TypeScript codebase for every app; ToolJet’s AI fills in the visual spec ToolJet’s builder has run since 2021: pages, components, queries and events.

What each produces. Superblocks’ eight-page property manager came to 81 TypeScript and CSS files, about 10,500 lines including the generated UI component library, with 55 npm packages declared (37 dependencies and 18 dev dependencies) and a 9,096-line lockfile. React 18.2.0 and Vite 6.4.3 are pinned to exact versions, and two packages are Superblocks’ own runtime. ToolJet’s builds of the same app are pages, components, queries and events in the visual builder, with no code files and no packages; our Claude Opus 5.5 build, for example, is 8 pages, 109 components, 17 queries and 109 events.

How you change it. You change a Clark-built app by prompting Clark again, which spends GAUs (you can point it at an element with Target), or by editing its React code in Superblocks’ Code view or your own IDE. Superblocks’ documentation describes no visual property editor for these apps. Everything ToolJet’s AI builds opens in the same visual builder people use by hand, so anyone on the team can select a component, query or event and change it without code.

Where the data lives. Our Superblocks Teams workspace had no writable database: its demo Postgres is read-only, so an app needs a database you connect, or keeps data in the browser or in memory, as our two Superblocks apps did. Superblocks also documents Superblocks Database, a managed Aurora PostgreSQL database in your own AWS account, set up through its Enterprise hybrid deployment; we did not test it. ToolJet Database comes with every ToolJet plan, and the AI creates the tables it needs.

What has to be maintained. Each Clark app carries its own pinned dependency list, so a security fix in one of those packages reaches an app only when that app’s dependencies are updated and it is published again. Superblocks helps here: it scans published apps’ npm packages for known vulnerabilities and offers a “Fix with Clark” flow that upgrades them, after which the app is redeployed. A fix to ToolJet’s shared runtime or components reaches every app with the platform: automatically on ToolJet Cloud, and with one upgrade on a self-hosted server. Custom code inside an app, and the databases and APIs it connects to, still need their own maintenance.

ToolJet lets you pick the model for every build; Clark runs on models Superblocks chooses.

Models. ToolJet’s picker offers 19 models from five providers, including GPT-6.1 Sol, Claude Opus 5.5 and Gemini 3.6 Flash, and shows each one’s typical build time and credit cost before you send a prompt. Clark has no model picker: it runs on models hosted by Superblocks, and pointing Clark at your own models (your Amazon Bedrock) is available only on Cloud-Prem deployments. Superblocks’ App AI setting picks the provider for AI features inside the apps you build, not for Clark itself.

Your own coding agent. ToolJet’s MCP plugin lets Claude Code or Codex build ToolJet apps directly, using your agent’s own subscription, so no ToolJet credits are spent, on every plan including Free. Superblocks offers two MCP servers. Its Builder MCP (in beta, by request) builds and edits apps from Claude, but Clark does the work in Superblocks, so the building still runs on Clark. Its Admin MCP lets coding agents list apps, manage integrations, deploy and manage access, without building apps. We did not test Superblocks’ MCP servers.

ToolJet Basic for two builders costs $46 a month and its credits cover about 190 eight-page builds on GPT-6 Luna; Superblocks’ entry pack of 100 GAUs costs $100 and covers about six.

The estimates below divide each plan’s monthly AI allowance by what our runs cost: the eight-page app alone, and the app plus both edits. They are estimates from single runs, on annual billing.

Plan	Price per month	AI allowance per month	Eight-page builds covered (estimate)	App plus two edits (estimate)	Published apps
ToolJet Basic, 2 builders (its maximum)	$46 (2 × $23)	4,000 credits	190 on GPT-6 Luna, 50 on GPT-6.1 Sol	104 on GPT-6 Luna, 19 on GPT-6.1 Sol	Up to 10
ToolJet Pro, 5 builders	$395 (5 × $79)	10,000 credits	476 on GPT-6 Luna, 127 on GPT-6.1 Sol	261 on GPT-6 Luna, 49 on GPT-6.1 Sol	Up to 10
ToolJet Team, 5 builders	$995 (5 × $199)	15,000 credits	714 on GPT-6 Luna, 190 on GPT-6.1 Sol	391 on GPT-6 Luna, 73 on GPT-6.1 Sol	Unlimited, with SSO and audit logs
Superblocks Teams, 100-GAU pack	$100	100 GAUs, shared by the organization	About 6	About 2	1 included, $10 a month for each extra app
Superblocks Teams, 500-GAU pack	$500	500 GAUs, shared by the organization	About 32	About 11	1 included, $10 a month for each extra app

Superblocks charges no per-builder fee: Teams is priced by the size of a shared GAU pool, from 100 to 2,500 GAUs a month, with top-ups when it runs low. ToolJet charges per builder, and each paid builder adds to the workspace’s shared credit pool; ToolJet credits are monthly with no daily cap. Prices here are annual billing; the per-run costs elsewhere use Superblocks’ monthly rate of $1.25 per GAU, and at the annual rate of $1.00 each Superblocks run costs a fifth less. All figures cover building and editing, not hosting or running the apps.

Sources: ToolJet pricing, Superblocks pricing, read on 5 October 2026.

Superblocks includes role-based access on Teams; ToolJet adds SSO and audit logs on its Team plan, where Superblocks keeps them for Enterprise.

	ToolJet	Superblocks
Role-based access	Admin, builder and end-user roles; custom groups on Team; page- and component-level permissions on Enterprise	Included on Teams
SSO	Team plan: Google, GitHub, OIDC, LDAP, SAML	Enterprise
Audit logs	Team plan	Enterprise
Self-hosting	Free open-source edition; paid plans self-host too	Hybrid or Cloud-Prem deployment on Enterprise
Source	Open source, AGPL-3.0	Proprietary

Sources: ToolJet pricing, ToolJet access control, Superblocks pricing, read on 5 October 2026.

For internal tools, pick ToolJet. It built the same apps as Superblocks for $0.38 instead of $53.71 for the eight-page app and its edits, stored every record in a database the whole team shares, passed every save check, has a free plan and a $23 entry plan, and keeps apps editable in a visual builder with no per-app codebase to maintain.

• Choose ToolJet when the app runs on company data many people use, when you want the model choice and cost visible per build, when non-developers need to edit what the AI built, when you want to build from your own coding agent without platform credits, or when you need to self-host without an Enterprise contract.
• Choose Superblocks when your developers want a React codebase as the source of truth, your data already lives in governed databases Superblocks connects to, and the per-run AI cost fits your budget.

Superblocks’ Clark was careful, finished every edit and was quicker at changes. For a team that will build and change many internal tools, the 26 to 140-fold cost difference, storage our tested apps kept in a browser or in memory, and the code-only editing matter more.

Explore ToolJet AI App Builder, compare ToolJet plans, or review the ToolJet vs Superblocks platform comparison for deployment and governance details.

Every build and edit in this article. ToolJet’s figures come from its credit meter; Superblocks’ from its per-app usage report.

Platform	Build	Model	Time	AI units used	Cost
ToolJet	Expense tracker	GPT-6 Luna	4m 18s	2.9 credits	$0.03
ToolJet	Expense tracker	GPT-6.1 Sol	2m 25s	22 credits	$0.22
ToolJet	Property manager	GPT-6 Luna	17m 24s	21 credits	$0.21
ToolJet	Property manager	GPT-6.1 Sol	11m 11s	78.55 credits	$0.79
ToolJet	Edit 1	GPT-6 Luna	10m 26s	8.1 credits	$0.08
ToolJet	Edit 2	GPT-6 Luna	11m 13s	9.2 credits	$0.09
ToolJet	Edit 1	GPT-6.1 Sol	8m 1s	64.5 credits	$0.65
ToolJet	Edit 2	GPT-6.1 Sol	7m 25s	61.17 credits	$0.61
ToolJet	Databricks dashboard	GPT-6 Luna	12m 49s	10.75 credits	$0.11
ToolJet	Databricks dashboard	GPT-6.1 Sol	9m 13s	87.44 credits	$0.87
Superblocks Teams	Expense tracker	Clark	About 4 min of building	7.40 GAUs	$9.25
Superblocks Teams	Property manager	Clark	About 13 min	15.52 GAUs	$19.40
Superblocks Teams	Edit 1	Clark	About 5 min	12.57 GAUs	$15.71
Superblocks Teams	Edit 2	Clark	About 4 min	14.88 GAUs	$18.60
Superblocks Teams	Databricks dashboard	Clark	About 9.5 min	9.36 GAUs	$11.70

ToolJet credits are $0.01 each; Superblocks GAUs are valued at $1.25, the monthly Teams price per GAU. The excluded first attempt at the Superblocks property manager used 14.10 GAUs and is not counted.

Every prompt below was sent word for word to both products. The Superblocks property-manager run reported here carried a stray word, “merg”, left in Superblocks’ prompt box from an earlier draft, at the end of the prompt.

Expense tracker:

Build an expense tracker. Let me add an expense with date, category (travel, meals, software, office, other), amount and a note. Show all expenses in a table I can filter by category and month, with the month's total at the top.
Design: a clean, modern light theme on a near-white background; a header with the app name and a short subtitle; the month's total as a large figure in a white card, with the number of expenses beside it; the add-expense form in its own card with the fields side by side and one primary button; category shown as coloured tags (travel blue, meals amber, software violet, office teal, other grey); indigo (#4F46E5) as the primary colour; money as $1,234.56 and dates as Mar 4, 2026.

Property manager:

Build a property management app for a small residential landlord, with eight pages and a left sidebar to move between them.
Seed realistic sample data: 4 properties, 32 units, 26 tenants, a lease for each occupied unit, 6 months of rent payments, 20 maintenance requests and 8 vendors.
1. Dashboard: KPI cards for occupancy %, rent collected this month vs expected, total overdue balance and open maintenance requests; a 6-month rent collection chart; occupancy by property as a bar chart; and a list of leases ending in the next 60 days.
2. Properties: a card grid showing each property's name, address, unit count and occupancy %; clicking a card opens a panel listing its units.
3. Units: a table with property, unit number, beds, baths, square feet, monthly rent and status (occupied, vacant, notice given); filters for property and status; rent is editable inline.
4. Tenants: a searchable directory; selecting a tenant shows their contact details, current lease, payment history and open maintenance requests.
5. Leases: a list with tenant, unit, start and end dates, rent, deposit and status (active, expiring, ended); a form to create a lease for a vacant unit, which marks the unit occupied; and a Renew action that extends the end date by 12 months.
6. Rent: a ledger of payments by month; a form to record a payment against a lease; and an overdue list showing days late, with a 5% late fee added once a payment is more than 5 days late.
7. Maintenance: requests grouped by status (new, scheduled, in progress, done) with priority (low, medium, high, urgent); a form to log a request for a unit; and the ability to assign a vendor and change status.
8. Vendors: a list with trade, phone, email and rating (1 to 5); the number of jobs assigned to each; and a form to add a vendor.
Design: a clean, modern light theme on a near-white background, white cards with soft borders, emerald (#059669) as the primary colour for the main buttons and active navigation, status tags in consistent colours (green for good states, amber for warnings, red for overdue or urgent, grey for neutral), money shown as $1,234 and dates as Mar 4, 2026.

Edit 1: “Add a ninth page, Reports, to the sidebar: a chart of monthly rent income against maintenance spend for each property over the last 6 months, a table with the same numbers per property and month, and a button to export that table as CSV. Track a cost on each maintenance request so the spend is real data, and fill costs in for the existing completed requests.”

Edit 2: “Change the primary colour from emerald to navy (#1E3A8A) everywhere in the app, show all money without cents (for example $1,234), and add a ‘Days to lease end’ column to the Leases list, sorted with the soonest first.”

Databricks dashboard (sent after connecting the workspace’s sample tables on each platform):

Build a sales analytics dashboard for Bakehouse, a bakery franchise chain, on my Databricks data. Use my connected Databricks data source. The tables are samples.bakehouse.sales_transactions, samples.bakehouse.sales_franchises and samples.bakehouse.sales_customers; read them to learn their columns. The data is read-only: query Databricks live, and do not create, copy or change any tables, and do not use sample or seeded data.
Three pages with a left sidebar:
1. Overview: a date range filter (defaulting to the full range in the data), a franchise country filter and a payment method filter that drive everything on the page. KPI cards for total revenue, transactions, units sold, average order value and active franchises, each with the change against the previous period of the same length. Charts: revenue by day as a line chart with a 7-day moving average; revenue by product as a horizontal bar chart showing each product's share; revenue by hour of day; revenue by day of week; payment method mix as a donut chart; and the top 10 franchises by revenue as a bar chart.
2. Franchises: a sortable table of every franchise with name, city, country, size, revenue, transactions, average order value and revenue rank, with country and size filters; selecting a franchise shows its revenue by day, its product mix and its busiest hours. Add a scatter chart of franchises by transactions against average order value.
3. Customers: KPI cards for unique customers, revenue per customer and repeat customers (more than one transaction); revenue by customer country; customers by gender; and a table of the top 25 customers by spend with name, country, transactions, total spend and favourite product.
Design: a clean, modern light theme on a near-white background; a header on each page with the page title and a one-line description; KPI cards in a row of white cards with soft borders, the value large and the change as a green or red tag; charts in white cards with a title and a short subtitle, in a two-column grid; amber (#D97706) as the primary colour for buttons, active navigation and the main chart series, with warm neutral colours for other series; money as $1,234 on KPIs and charts and $1,234.56 in tables, and dates as Mar 4, 2026.

• Superblocks pricing
• Superblocks: developing apps in code
• Superblocks: bring your own inference
• Superblocks: Builder MCP
• Superblocks: Admin MCP server
• ToolJet pricing
• ToolJet access control
• ToolJet MCP

All read on 5 October 2026.