Softr Review 2026: Pricing, Users & Portal Limits

The current pricing page advertises unlimited apps, databases, workflows, agents, and forms on Free. However, Free includes only five combined app users, up to three builders, 500 monthly workflow actions, and five monthly AI credits. Database record and structural ceilings also apply. Creating another app therefore does not manufacture another independent allowance for users, automation, or AI usage within the same workspace.

Free does not include a custom domain. The plan guide gives every paid plan one domain, with additional domains costing $13/month on annual billing or $15 monthly. Domain registration remains a separate consideration. Basic includes one builder, compared with Free's three; extra Basic builders cost $5/month annually or $6 monthly. Upgrading can change the shape of the allowance, not simply increase everything.

A premium trial is separate from permanent Free access. The current guide describes seven initial days and another seven after attaching a card, with billing after the trial unless canceled. Some older FAQ wording differs, so verify the offer shown during signup rather than assuming a month of unrestricted access. Use the no-code category guide to compare permanent capabilities, not features temporarily unlocked during evaluation.

The official rate table lists Basic at $19/month annually or $25 monthly, Pro at $99 or $119, and Business at $329 or $395. Enterprise is custom-priced. Annual billing pays for twelve months upfront: the listed base equivalents produce $228, $1,188, and $3,948 respectively, before tax and add-ons. Older Professional price grids should not replace these current plan names and amounts.

Users are counted by unique email across all published apps in a workspace. The same email in three apps counts once, deactivated users do not count, and builders can use apps without consuming app-user capacity. Team users share a business email domain with a workspace builder; everyone else is a Client user. Free email domains such as Gmail count as Client identities, regardless of someone's actual employment relationship.

Basic includes five Team and five Client users, with no extra-user add-on. Pro includes ten Team and fifty Client users, expandable to fifty and 250. Business includes thirty and 100, expandable to 100 and 500. Pro and Business include two and three builders respectively. These separate pools matter: spare Client capacity is not documented as interchangeable with Team capacity, and builder access is not a harmless billing workaround.

Extra Pro users cost $3 per Team user and $1 per Client user each month; Business extras cost $5 and $2. Extra-user billing remains monthly even with an annual base. For illustration, sixteen Team users and eighty Client users on annual Pro add $48 per full month to the $1,188 annual base. Daily checks, the documented initial seven-day grace period, and proration affect real invoices; decreasing usage does not create an immediate mid-cycle refund.

The database documentation specifies 1,000 records per database and 5,000 across databases on Free. Those are different limits, not alternative descriptions of the same capacity. A single 2,000-record project would not fit the stated per-database ceiling simply because the workspace total remains below 5,000. When comparing Airtable, map the actual relational dataset rather than just counting visible rows on one screen.

Paid pricing cards show 50,000 records for Basic, 500,000 for Pro, and one million for Business. They do not establish that every newly created database receives that entire allowance independently. Exact paid per-database ceilings were not resolved by the reviewed material and should be confirmed before a large import. Native record capacity also should not be substituted for an external provider's storage, API, or subscription limits.

The database guide additionally lists structural ceilings of 1,000 databases per workspace, 512 tables per database, and 512 fields per table. That qualifies the marketing shorthand of unlimited databases. The Airtable alternatives guide helps separate a database replacement from an interface layered over data you intend to keep elsewhere.

Softr's permissions overview separates page and block visibility from record access and permitted actions. The default groups distinguish logged-in from non-logged-in visitors. Current plan documentation assigns three custom groups to Pro and ten to Business. Team/Client billing classifications do not automatically create these business roles: a paid Client user still needs rules defining which client's information they may see.

Business adds global data restrictions, applying record rules across an application rather than relying only on repeated block configuration. Pro's action visibility and branding controls are therefore not equivalent to the complete Business permission package. The overview still marks field-level global restrictions as coming soon; do not budget around that as an available feature. Enterprise identity capabilities such as SAML/OpenID SSO and SCIM are a separate purchasing decision.

Before launch, evaluate two synthetic clients, an internal editor, and a logged-out visitor. Check direct record URLs, edits, downloads, and any workflow that writes through a connected account. Hiding a menu item is not sufficient evidence that another client's record cannot be retrieved. Apply the same permission-first evaluation when comparing Bubble.

The Workflows guide documents app actions, database events, schedules, external triggers, and connected services. Native workflows may cover a portal's approval and notification path without a separate automation product. Nevertheless, compare the actual integrations and failure handling with Zapier and Make before replacing an existing process.

Current monthly allowances are 500 actions on Free, 2,500 on Basic, 10,000 on Pro, and 25,000 on Business. Usage counts executed actions, not merely the number of workflows created. Triggers and flow controls such as Wait, Branch, Filter, and Bulk actions are excluded; some integration actions can consume more than one execution. An AI step may also draw from the separate AI-credit pool.

As a planning example, 200 submissions executing three ordinary single-unit actions each consume 600 actions, even though there are only 200 runs. Check the actual step types before applying that arithmetic. Unused actions do not roll over, and the pricing FAQ says workflows pause at the limit until renewal or upgrade. Define who notices exhaustion and handles unfinished business requests instead of assuming a paid upgrade will automatically repair missed work.

The AI-credit guide lists five monthly credits on Free, ten on Basic, fifty on Pro, and 100 on Business. All AI features draw from the same pool. Consumption depends on model, task, and input/output size, so fifty credits do not establish fifty completed apps, fifty records, or fifty accurate answers. Database agents and workflow AI should be budgeted alongside other enabled AI features.

Included credits reset monthly without rollover. Paid-plan credit add-ons are recurring subscriptions rather than one-off purchases, and unused add-on credits can roll over for up to one month. When credits run out, AI features pause; the guide does not describe an automatic surprise overage charge. The pricing page separately confirms that the visual editor remains usable without AI, which is an important distinction from an AI-only development dependency.

Compare Lovable if prompting is central to the build process, but keep app generation, runtime AI, and authorization as different questions. Before enabling an assistant on customer records, identify which connected services process the inputs and approve that data use. Generated classifications and summaries need a review path appropriate to their consequences.

The official export guide documents CSV and XLSX downloads from Table blocks. PDF export applies to the newer Item Details block and can capture that block or the page. These are useful data and report outputs, but their existence does not prove that every attachment, hidden field, relationship, user identity, and workflow is included in a recoverable application archive.

The reviewed documentation did not establish a complete source-code or self-hostable app export. Treat that as an unresolved portability requirement, not a promise that a CSV recreates Softr elsewhere. Keeping records in another service can simplify access to those records, but the interface, authentication, permissions, and automations still need an exit plan. Compare the exact handoff requirements with Lovable rather than equating all app-building approaches.

For an authorized sample, check field types, linked records, file downloads, and whether a different application can use the resulting dataset. Record business rules separately and protect exported customer information with the same care as live records. A documentation-oriented destination such as Notion may accept some information without reproducing the original portal.

Softr is a credible shortlist option for an operations team that values a managed portal, native data, and connected workflows more than ownership of a fully portable application package. The decision should be based on the audience and access model first. Five public-facing pages are not necessarily a larger project than one page holding confidential client records; the latter may require a materially more expensive permission tier.

Choose Free for a bounded evaluation, Basic for a genuinely small tool without custom-group requirements, and assess Pro or Business against the controls the portal actually needs. Revisit the no-code and automation category if the projected identity bill, integration requirements, or migration obligations outweigh the convenience.

Before paying annually, obtain a clear user-count breakdown, confirm the applicable record ceiling, and evaluate a complete representative journey from authorized login to exported data. Assign an owner for usage alerts and access changes.