Best No-Code AI

What no-code AI costs: plans, credits and the bill for a small company

Entry prices cluster between $20 and $40 a month. The real bill depends on three numbers each tool counts differently - people, records and AI usage.

By The editors · Updated · 3 min read

A finance manager with a grey beard reviews a stack of printed monthly invoices with a pen and a desk calculator, a mug beside them.
Subscriptions add up one invoice at a time.
Contents
  1. What each tool counts
  2. A worked example: a 20-person company
  3. The costs that do not appear on pricing pages
  4. How to budget the first year
  5. Questions buyers ask
  6. Sources

No-code AI looks cheap at the entry level, and often is. The difficulty is that each tool counts something different (editors, tasks, operations, workload units or credits) and AI is usually metered on top. Stanford's AI Index reports that the cost of running a model at GPT-3.5's level fell more than 280-fold between November 2022 and October 2024⁠1; much of that saving reaches buyers as cheaper credits rather than flat fees.

What each tool counts

  • Editors. Airtable charges per collaborator with edit rights: Team $20 a month billed annually, Business $45. Read-only users and form submissions are free⁠2. AI credits come per paid user, 15,000 a month on Team⁠3.
  • Tasks. Zapier charges by monthly task volume: Professional $19.99 a month billed annually for 750 tasks, Team $69 for 2,000; agent activity is counted separately⁠4.
  • Credits per operation. Make counts one credit per operation by default, plus token-based credits for AI on its own provider⁠56.
  • Credits as money. Gumloop indexes credits at $0.005 and adds an 8% orchestration fee to list costs; Pro starts at $37 a month for 20,000 credits⁠7.
  • Credits as tokens. Glide's credits are standardised model tokens; Basic costs $25 a month, billed monthly⁠8.
  • Workload units. Bubble meters server work; Starter is $32 a month billed annually for 175,000 units, with overage beyond⁠9.
  • A plan plus hosting. OpenLazarus publishes Starter at $100 and Pro at $300 a month, each including that amount in AI credits, plus managed hosting from $150⁠1011.

A worked example: a 20-person company

Take a 20-person services company with five people who edit records daily and fifteen who mostly read. It wants a shared database, automations between its existing tools and AI on routine records. Using published list prices (our arithmetic, before taxes, add-ons and AI overage):

ApproachTools and plansMonthly cost
LayeredAirtable Team, 5 editors ($100)⁠2 + Zapier Team ($69)⁠4 + Gumloop Pro ($37)⁠7about $206
One systemOpenLazarus Starter ($100) + hosting from $150⁠1011from $250

The layered stack is cheaper on paper and keeps the company's current CRM, invoicing and mail tools, which continue to be paid for separately. The single system costs a little more and is designed to replace some of those subscriptions; whether it does depends on which tools the design actually retires. In both cases, AI usage above the included allowance is extra.

The costs that do not appear on pricing pages

  • Overage and caps. Bubble charges overage once workload units pass 100% unless it is turned off⁠9. Gumloop's agents stop when credits run out until pay-as-you-go is enabled⁠7.
  • Expiring credits. Make's extra credits cost 25% more than plan credits and expire⁠12; Airtable's AI credits do not roll over⁠3.
  • Governance upgrades. Audit logs and SSO often require a higher plan, such as Zapier Team or Make Enterprise⁠413.
  • Glue. Every layer added to a stack is another set of connections and permissions for someone to maintain.
  • Upkeep. A person's time, or a paid support plan such as OpenLazarus developer support from $4,500 a month⁠11.

How to budget the first year

  1. Count three numbers: people who edit, records you will hold in a year, and AI actions a month.
  2. Price the plan that includes the governance you need (SSO, audit logs, approvals), not the entry plan.
  3. Run a two-week pilot and read the usage report before committing to annual billing.
  4. List the subscriptions the new tool replaces, and subtract them only once they are cancelled.
  5. Set alerts or caps wherever a tool offers them, and decide in advance what should happen when a cap is reached.

A sound budget ends the year with no surprise upgrades, a known cost per process and at least one subscription retired.

Questions buyers ask

Is no-code AI cheaper than hiring a developer?

For a first internal system, usually yes. Entry plans cost tens of dollars a month. The comparison changes as usage grows, because AI is metered separately, and because someone still has to maintain the system. A support plan or a person's time belongs in the comparison.

Why are AI credits so hard to budget?

Because one credit buys different amounts of work. At Airtable a record categorisation costs 1 credit and a document analysis 200; at Make an agent costs one credit per operation plus tokens. Run a short pilot and measure.

Which no-code AI tool has the most predictable price?

Tools priced per seat or per plan are easiest to forecast. Airtable prices per editor, Zapier by task tier, and OpenLazarus by a plan that includes AI credits plus hosting. Credit-only models are the hardest to predict.

Sources

  1. The 2025 AI Index Report. Stanford HAI. Accessed 8 Nov 2026.
  2. Airtable plans overview. Airtable Help Center. Accessed 8 Nov 2026.
  3. Airtable AI billing. Airtable Help Center. Accessed 8 Nov 2026.
  4. Zapier pricing. Zapier. Accessed 8 Nov 2026.
  5. Credits. Make Help Center. Accessed 8 Nov 2026.
  6. Credit usage for AI agents. Make Help Center, 10 Jun 2026. Accessed 8 Nov 2026.
  7. Pricing. Gumloop. Accessed 8 Nov 2026.
  8. Pricing: Free, Pro and Enterprise plans. Glide. Accessed 8 Nov 2026.
  9. Bubble pricing. Bubble. Accessed 8 Nov 2026.
  10. Questions about OpenLazarus (FAQ). OpenLazarus. Accessed 8 Nov 2026.
  11. Pricing. OpenLazarus. Accessed 8 Nov 2026.
  12. Adjustments to plans and pricing. Make Help Center, 6 Nov 2025. Accessed 8 Nov 2026.
  13. Audit logs. Make Help Center. Accessed 8 Nov 2026.

Updated 8 Nov 2026: First published.

Further reading