A8gent
AI Agent PlatformsAI agent builder

Lindy Pricing and Review (2026)

AI assistants that automate your busywork.

Reviewed by A8gent
Contents
  1. Verdict
  2. What is Lindy?
  3. Best for, and not for
  4. Strengths and limits
  5. Pricing
  6. What people actually build with it
  7. What Lindy actually costs at small-business volume
  8. The gotcha: overages cost double and there is no cheaper way to buy
  9. What breaks in production, and how you find out
  10. What it connects to that you already run
  11. Lindy alternatives: what to pick instead, and when
  12. Migration and lock-in: can you get your work out?
  13. FAQ
VerdictIs Lindy worth it?

Lindy is an AI assistant priced like a subscription and metered like a utility, and the two do not quite meet in the middle. As of August 2026 the plans are Plus at $49.99 a month with two connected inboxes, Pro at $99.99 with three inboxes plus computer use and model selection, Max at $199.99 with five inboxes, and a custom Enterprise tier with HIPAA, SSO, SCIM and audit logs. Usage is described only as a multiplier: Pro is three times Plus, Max is seven times. The docs do quantify credits for custom agents at one to three per task on basic models and around ten on large ones, but no plan states how many credits it includes. That single omission is the review.

What is Lindy?

Lindy is a no-code platform for building AI assistants that automate tasks like email handling, meeting scheduling, CRM updates, and lead follow-up. You describe what you want or start from a template, and Lindy sets up an agent that triggers on events and acts across connected tools. It targets everyday business workflows, especially around email and customer-facing tasks.

Best for

Individuals and small teams who want AI assistants for email, scheduling, and CRM busywork.

Not for

High-volume customer-facing support. Lindy is an assistant for your inbox and calendar, not a widget on your website, and pricing 6,000 monthly agent responses against an unpublished credit allowance is guesswork.

Strengths

  • Genuinely good at the assistant job: inbox triage and labelling, drafting replies in your voice, meeting scheduling, prep, recording and follow-up
  • The credits documentation is unusually candid about what drives cost and gives four concrete optimisation levers, including clearing task context after large data steps
  • Per-step credit visibility: open a task, hover a step, see what that step consumed
  • Model selection on Pro and above lets you trade cost against capability yourself, which is the single biggest lever on AI spend
  • Computer use on Pro and above reaches systems with no API at all, slowly but sometimes uniquely
  • A usage warning pauses resource-heavy tasks and asks before continuing, which is a genuinely good default for an agent with your credit card behind it
  • Enterprise carries HIPAA with a signed BAA, SSO, SCIM and audit logs, which clears a real bar for regulated practices

Limitations

  • No plan states how many credits it includes, so you cannot model whether your workload fits before you run it
  • Usage is expressed as a multiplier of an unstated base, which is not a unit you can budget against
  • Tiers are gated on connected inbox count, an odd constraint if you have many mailboxes and light usage
  • Overages bill at twice the standard credit rate, and there is no option to buy one-off credits at the normal price
  • When you hit the limit the assistant pauses until the next cycle and paused tasks do not restart automatically; you restart them by hand
  • Credits do not roll over and the cycle runs from your signup date rather than the calendar month
  • No free tier at all, only a seven-day trial, so evaluation runs on a clock
  • Cloud only, so data residency requirements rule it out below Enterprise

Lindy pricing

Lindy sells flat monthly plans differentiated by connected inbox count and a relative usage multiplier, then meters the work in credits underneath. Credits are consumed whenever an agent runs a task, where a task is everything the agent does after being triggered, including conditions, actions, agent steps and messages. The docs put most tasks at one to three credits on basic models and around ten on large models, with a one-credit minimum per task, and list four cost drivers: model intelligence, task complexity, premium actions and how long the agent runs. Credits do not roll over. When you run out, agents pause and new tasks are held. There is no option to buy one-off credits: you either upgrade to a higher plan month-to-month or toggle on overages, which bill at twice the standard credit rate. There is no free tier, only a seven-day trial with full Pro features.

Lindy pricing tiers
PlanPriceWhat you get
Plus$49.99/mo2 connected inboxes, standard usage, inbox triage, drafts, meeting scheduling, recording and notes, hundreds of integrations (as of Aug 2026)
Pro$99.99/mo3 inboxes, 3x Plus usage, computer use, model selection, 1:1 onboarding session
Max$199.99/mo5 inboxes, 7x Plus usage. The highest published price Lindy offers
EnterpriseCustomShared usage plus bonus credits, HIPAA with signed BAA, SSO and SCIM, custom company context, audit logs, dedicated support
Credit rates1-10 creditsMost tasks 1-3 credits on basic models, around 10 on large models, minimum 1 credit per task. Included allowance per plan is not published
Overages2x standard rateOpt-in toggle in billing settings. Usage beyond the plan limit bills at twice the normal credit cost; one-off credit purchases are not offered
Free trial7 daysFull Pro features, cancel before it ends to avoid charge. There is no permanently free tier

Pricing reflects public plans as of August 6, 2026 and can change. Check Lindy for the latest.

What people actually build with it

Lindy is shaped like a person rather than a pipeline, and the builds that work reflect that. The trigger is usually an email, a calendar event or a message, and the outcome is usually a draft, a booking or a summary that a human then approves.

Inbox triage and drafted replies

The flagship use case: Lindy watches a connected mailbox, labels and prioritises incoming mail, and drafts replies in your voice for the ones that need answering. For a solo consultant or a partner in an accounting firm, that is an hour a day of low-value sorting removed without changing any system.

The credits documentation makes an important point here that applies to every trigger-based build: filter the trigger. If you want Lindy to label mail from one client, filter on the sending domain so the agent does not run on every message that arrives. Every run is at minimum one credit, and an unfiltered inbox trigger is the commonest way to burn an allowance you cannot see.

The whole meeting lifecycle

Scheduling, prep briefings, recording and notes are bundled on every plan including Plus, and this is where Lindy is genuinely differentiated. The realistic outcome is that meetings arrive with a one-paragraph brief on who you are meeting and what happened last time, and leave with notes and follow-ups drafted. The docs rate meeting recording and notes as moderate resource use and everything else in that lifecycle as low or very low.

Ad hoc research, with a warning attached

You can delegate open-ended research tasks, and Lindy's own usage table marks ad hoc research as the one category whose resource impact varies. This is also the case that most often trips the usage warning, where Lindy pauses mid-task because it is consuming more than expected and asks whether to continue. Treat that pause as a feature: it is the difference between a research task costing what you expected and a research task quietly costing a plan tier.

The one it is bad at

Customer-facing support at volume. Lindy has no website widget story, its pricing is gated on inbox count, and the reference workload used across this batch - 1,000 conversations a month at six turns - is roughly 6,000 triggered tasks. At the documented one to three credits per task on basic models, that is 6,000 to 18,000 credits against an allowance nobody will tell you. For that job, look at Chatbase or Botpress instead, and treat Lindy as the assistant behind your own team's inbox rather than the front door for customers.

Bottom line: Lindy earns its money on inbox and meeting work for individual professionals, and is the wrong shape for customer-facing support at volume.

What Lindy actually costs at small-business volume

The subscription side is simple and honest. Plus is $49.99 a month, Pro $99.99, Max $199.99, each billed monthly with no long-term contract on Plus and cancellation from account settings. Lindy's own pricing page frames those numbers against the cost of a human assistant, which is the right comparison and a defensible one.

The usage side is where forecasting fails. A task is everything an agent does after a trigger fires: every condition, action, agent step and message in that run. The docs put most tasks at one to three credits on basic models and around ten on large ones, with a floor of one credit. What no published page states is how many credits Plus includes. Pro is three times Plus and Max is seven times Plus, but three times an unpublished number is still an unpublished number.

So on our cross-batch reference workload, Lindy is the only one of the five we cannot price. The table below shows why the comparison breaks down, and what you can say honestly instead.

The same workload - 1,000 support conversations a month at six turns - expressed in each platform's unit, as of August 2026. Lindy is the only entry that cannot be converted to a monthly figure from published information.
PlatformUnitWorkload in that unitMonthly cost
LindyCredit per task, 1-3 basic / ~10 large~6,000 tasks, so 6,000-60,000 creditsNot publishable: no plan states its credit allowance
ChatbaseMessage credit per response6,000-12,000 credits$200-$400
BotpressConversation1,000 conversations$670-$750 billed annually
VoiceflowCredit, 200 to the dollar~11,250 credits~$56 usage plus unpublished plan fee
MindStudioTokens at provider cost plus $20 seat~16M tokens~$26

What you can budget with confidence

For the job Lindy is actually built for, the numbers behave. One professional's inbox and calendar is a low-resource workload by Lindy's own table: triage very low, drafting low, scheduling very low, meeting notes moderate. Most users, the docs say, never come close to their plan's limits. For a single user on Plus at $49.99 that is a predictable bill and a fair one.

The forecast breaks when volume or ambition rises: several mailboxes, heavy ad hoc research, or a custom agent processing hundreds of rows at a time. Those are precisely the workloads where you need the allowance published, and precisely where it is not.

Bottom line: Lindy's subscription is predictable and its usage is not: at $49.99 to $199.99 a month the plan price is clear, but no published page states how many credits any plan includes.

The gotcha: overages cost double and there is no cheaper way to buy

This is the detail that should change decisions and it sits in the docs rather than the pricing page. If you exceed your plan's usage limit, overage usage is charged at twice your plan's standard credit rate. And when you ask the obvious follow-up - can I just buy more credits at the normal price - the answer is explicit: there is no option to purchase one-off credits. Your two choices are to upgrade to a higher plan month-to-month, or to toggle overages on and pay 2x.

Combine that with the fact that the allowance is not published, and you get a structural problem. You cannot tell in advance which side of the line you will land on, and the penalty for landing on the wrong side is a doubled rate on everything past it. The rational response is to run a representative fortnight, watch the credit gauge at the bottom left of the workspace, and choose the tier from observed consumption rather than from the plan descriptions.

The second gotcha is what happens if you do nothing. With overages off, your assistant pauses until the next billing cycle. Your data and settings are safe, but paused tasks do not restart automatically when the plan resets - you restart them by hand. For an assistant handling your inbox, a pause you did not notice is a week of unanswered mail, and the reset date is your signup anniversary rather than the first of the month, which is exactly the kind of date nobody remembers.

Bottom line: Lindy bills overage at 2x the standard credit rate and sells no one-off credits, so the only cheap way to handle a heavy month is to upgrade before it happens.

What breaks in production, and how you find out

An agent with permission to act inside your mailbox has a different risk profile from a chatbot on a website, and the failures follow from that.

Sending is irreversible

The failure that matters most is an agent sending something wrong to someone real. Unlike a failed database write, you cannot undo it, and unlike a bad chatbot answer, the recipient did not choose to start the conversation. Keep drafting and sending separate for anything customer-facing until you have watched it for a month, and restrict which addresses it may write to. A human approval step is not a training-wheels measure here; it is the control, and keeping a human in the loop on outbound mail costs you a few seconds against an apology you cannot take back.

The related failure is context confusion across threads: a reply that is entirely correct, attached to the wrong conversation. It reads as competence, which is what makes it dangerous. Sample sent mail weekly rather than waiting for a complaint.

Cost surprises inside a single task

Because a task is everything after the trigger, a single run can be arbitrarily expensive. The docs are explicit that agents run until completion and that carrying conversation history forward between steps drives usage up, especially with large data like webpage contents or spreadsheets. Their fix is a Clear Task Context action after any significant data-processing step, and it is worth building in from the start rather than after the bill.

Lindy's usage warning is the safety net: when a task starts using significantly more resources than usual it pauses and asks whether to continue. You can let it finish or stop it, and stopping consumes nothing further. Treat every one of those warnings as a signal to go and look at the task rather than as a dialog to dismiss.

The quiet stop

The third pattern is the assistant simply stopping. Credits run out, agents pause, new tasks are held, and nothing dramatic happens on screen. The credit gauge in the bottom left is the only routine warning, and the usage section under Settings then Billing is the only place to check. Put a monthly reminder in your calendar three days before your signup anniversary and look at the gauge. It is a crude control, and on a platform that does not publish allowances it is the one that works.

Bottom line: The three Lindy failure modes are an irreversible send, a single task that consumes far more than expected, and a silent pause when credits run out; only the middle one warns you by itself.

What it connects to that you already run

Model choice is itself a connection worth naming, because on Pro and above you pick which model powers your Lindy, and the practical spread between a cheap small model and a frontier one like Claude is roughly the difference between one credit a task and ten. Lindy connects to hundreds of services, and the ones that matter are the ones closest to its job: Gmail and Google Calendar, Microsoft Outlook for mail and calendar, Slack, Notion, HubSpot and Salesforce. Multi-account setup lets it work across several Google or Outlook accounts, which is precisely what the inbox-count gating on each plan is metering.

Beyond the catalogue there are two escape hatches. Computer use, on Pro and above, lets Lindy operate web applications on your behalf, which reaches systems that have no API at all - slower and more fragile than an integration, and sometimes the only route into an old practice management or booking system. And for scraping and data work, the skills list includes Apify actors, which is a more robust answer than pointing computer use at a website.

What Lindy is not is an integration platform. If the requirement is moving records between two systems reliably on a schedule, that is a job for Zapier, Make.com or n8n, and plenty of businesses sensibly run both: an assistant for the human-facing coordination and an automation platform for the machinery behind it.

One integration detail is worth planning around because it interacts with the pricing. Each connected inbox is both a capability and a cost gate: Plus allows two, Pro three and Max five. A business with six shared mailboxes, each lightly used, is pushed up the tiers by mailbox count rather than by work volume, which is the wrong incentive. The workaround most teams land on is consolidating aliases into fewer real mailboxes before connecting them, which is worth doing anyway.

There is no self-hosted option at any published tier, so if a data residency requirement rules out a cloud assistant handling your mail, the conversation starts at Enterprise. That is also the only tier carrying HIPAA with a signed BAA, SSO, SCIM and audit logs, so a regulated practice should treat Max as a ceiling for evaluation rather than a plan it could stay on.

Bottom line: Lindy's integration strength is depth on mail and calendar rather than breadth, with computer use as the escape hatch for systems that have no API.

Lindy alternatives: what to pick instead, and when

Lindy competes in two directions at once. Against other assistants it competes on how well it handles an inbox. Against automation platforms it competes on whether you wanted an assistant at all. Most people shopping for Lindy should check the second question first, because the answer changes the budget by an order of magnitude.

Lindy against the tools most often compared with it, priced where a comparison is possible, as of August 2026.
AlternativeBilling unitCost shapePick it instead when
ZapierTaskFree tier, then from $29.99/moThe work is a repeatable sequence between systems, not judgement
n8nWorkflow executionSelf-hostable, low and predictableYou want to own the infrastructure and cap the cost yourself
MindStudioFlat seat plus tokens at cost~$20/mo plus usageYou want to see exactly what the model spend is, per run
ChatbaseMessage credit$40/mo monthly-billed upwardThe job is answering customers, not assisting you
Copilot StudioMessage and credit packsTenant credit packsYou are a Microsoft 365 shop and the assistant should live there

When an automation platform is the honest answer

If your process is invoice chasing, syncing stock levels, or moving records between a CRM and a spreadsheet, an assistant is a strange and expensive way to do it. A workflow tool costs a fraction, fails more predictably and shows you exactly which step broke. Lindy's own strength - open-ended judgement across a mailbox - is the thing you are not using in that scenario.

The reverse also holds. If your problem genuinely is that too much coordination lands in one person's inbox, no amount of workflow automation fixes it, because the work is judgement rather than routing. That is the case where Lindy's $49.99 to $199.99 a month is easy to justify against the alternative of another salary.

When credit transparency is the deciding factor

If the unpublished allowance is a dealbreaker for your finance process, MindStudio is the sharpest contrast in this batch: a flat plan plus model usage passed through at provider cost, with a published rate for every model, a workspace budget you can set, and per-user monthly limits that suspend usage rather than billing double. It is a builder rather than an assistant, so you are trading polish for visibility, but you will always know what a run cost.

Bottom line: Pick a workflow tool if the task is a repeatable sequence, pick MindStudio if unpublished credit allowances are a dealbreaker, and pick Lindy only if the work is genuinely assistant-shaped.

Migration and lock-in: can you get your work out?

Lindy has one of the softer lock-in positions in this batch, and it is a real advantage of assistant products. There is no large body of workflow logic to migrate. What accumulates is configuration, instructions, memories and context about how your business works, plus whatever the agent has learned about your writing style.

That accumulated context is the actual asset and it is worth writing down separately from day one: who your important contacts are, what you always say yes to, what must always come to you, how you want a decline phrased. Kept as a document, it transfers to any assistant and survives you changing your mind about vendors. Kept only inside Lindy's memories, it is a thing you would rebuild.

The commercial exit is clean. Plus carries no long-term contract, you cancel from account settings and keep access to the end of the billing period. Plans change immediately in either direction. There is no annual lock at the published tiers, which also means there is no annual discount to weigh.

Two things do bind. Connected inbox permissions are yours to revoke, but re-establishing an equivalent setup elsewhere means reconnecting mail, calendar and every integration you added, which is a day of work rather than an export. And on Enterprise you are into a negotiated agreement with shared usage and bonus credits, so the renewal date becomes your exit window in the usual way.

Bottom line: There is little to migrate and much to write down: keep the instructions and context that Lindy has learned in a document outside the platform, because that is the only asset worth months.

Lindy FAQ

How much does Lindy cost?

As of August 2026, Plus is $49.99 a month with two connected inboxes, Pro is $99.99 with three inboxes plus computer use and model selection, and Max is $199.99 with five inboxes. Enterprise is custom quoted and adds HIPAA with a signed BAA, SSO, SCIM, custom company context and audit logs. Prices are monthly with no long-term contract on Plus, and there is no annual discount at the published tiers. A seven-day trial with full Pro features is the only free access.

What is a Lindy credit and how many do I get?

A credit is consumed when an agent runs a task, where a task is everything the agent does after its trigger fires: conditions, actions, agent steps and messages. Lindy's docs put most tasks at one to three credits on basic models and around ten on large models, with a minimum of one credit per task. How many credits each plan includes is not published anywhere. The plans describe usage relatively - Pro is three times Plus, Max is seven times - which is not a number you can budget against.

Does Lindy have a free plan?

No. There is a seven-day free trial with full Pro features and you can cancel before it ends without being charged, but there is no permanently free tier. That matters for evaluation: with an unpublished credit allowance, the only reliable way to size a plan is to run a representative period and watch the credit gauge, and seven days is short for a workload that varies week to week. Plan the trial around your busiest week rather than your quietest.

What happens if I hit my Lindy usage limit?

Your assistant pauses until the next billing cycle, which runs monthly from your signup date rather than the calendar month. Data and settings are safe, but paused tasks do not restart automatically - you restart them by hand. Two ways out exist: upgrade to a higher plan month-to-month, or toggle on overages in billing settings, which lets work continue at twice the standard credit rate. Lindy does not sell one-off credits at the normal price.

Why is Lindy overage twice the price?

That is Lindy's stated policy: overage usage is charged at 2x your plan's standard rate. Structurally it is a nudge to upgrade rather than to overspend, and it works, because upgrading month-to-month is always cheaper per credit than paying double. The practical implication is that you should size the plan from observed usage rather than optimism. Run a representative fortnight, read the credit gauge at the bottom left of the workspace, and pick the tier that covers your busy weeks, not your average one.

Lindy vs Zapier: which should we use?

Ask what the work looks like. If it is a repeatable sequence of steps between systems - a form fills, a record is created, a notification goes out - Zapier is cheaper, more predictable and easier to debug, and Lindy is the wrong shape. If the work is judgement across a mailbox and a calendar, triaging, drafting and coordinating, Zapier cannot do it and Lindy is purpose-built. Plenty of businesses run both, and the split usually falls neatly along that line.

Can Lindy handle customer support?

Not at volume, and not in the way people usually mean. Lindy has no website chat widget, its plans are gated on connected inbox count, and our reference workload of 1,000 conversations a month at six turns is roughly 6,000 triggered tasks, which at one to three credits each is a number you cannot check against an unpublished allowance. For customer-facing support look at Chatbase for documentation-grounded Q&A or Botpress if you need designed flows, and keep Lindy for your own team's inbox.

How do we reduce Lindy credit usage?

Lindy's own documentation gives four levers and they are the right ones. Filter high-volume triggers so agents only run on relevant events, because every run costs at least one credit. Use a cheaper model for straightforward work like summarising or routing. Add a Clear Task Context action after any large data-processing step, since carried-forward history inflates every subsequent step. And compare actions, because some cost more than others for the same outcome.

Can you set up the automation side for us?

Yes, and pairing the two is usually the right answer: an assistant for inbox and calendar coordination, a proper automation for the repeatable back-office processes behind it. We scope which is which on a free automation audit, build the automation around the tools you already run, and hand it over with documentation you own plus a recurring plain-language report on what ran, what it saved and what needs attention.

Looking at alternatives? Lindy is better than us at being an assistant. If what you want is something that lives in your inbox and calendar and handles the coordination a person would, it is purpose-built and its pricing against a human assistant is a fair framing. Two honest criticisms stand: it does not publish what a plan's credits buy, and it charges double for overage while selling no one-off credits, which together make budgeting harder than it should be. Autonoly is aimed at a different job - recurring operational work that runs on a schedule with nobody watching - and plenty of businesses should run both. What we would not do is use an inbox assistant for customer-facing support at volume, and neither should you. See the Autonoly review.