Zapier pricing

About

Zapier is the workflow-automation platform that connects 9,000+ apps so non-developers can move data between tools and automate business processes without writing code. Founded in 2011, it pioneered the “Zap” — a trigger-plus-action automation — and has since expanded into a broader “AI orchestration” platform that now bundles Zaps (workflows), Tables (structured data), Forms (custom intake), and Zapier MCP (an AI action layer connecting LLMs to those 9,000+ apps) into a single plan. Zapier is a profitable, bootstrapped-to-late-stage private company widely reported to have crossed $300M+ in annual recurring revenue, making it one of the largest independent SaaS automation vendors.

The company serves a wide spectrum, from solo operators and small teams on self-serve plans to Fortune 500 enterprises buying governed, org-wide deployments. Its 2026 positioning leans hard into AI: the marketing frames Zapier as “the orchestration layer that makes AI safe,” emphasizing governance (action restrictions, BYOM, audit logs, log streaming to Datadog/Splunk) as the differentiator against a field of point AI tools.

Zapier competes with no-code automation peers (Make, n8n, Workato) on the platform side and increasingly with AI-agent builders on its newer add-on products. Its moat is breadth of integrations plus a 13-year-old production-grade execution infrastructure that handles retries, token refresh, and credential storage so AI agents don’t have to.

Pricing summary : How Zapier’s task-metered platform plus AI add-ons work

Zapier’s core Platform is usage-based, metered by tasks — a task is any successful action that runs in Zapier. As of June 2026, Zapier unified its meter: Zap workflows, AI steps, Code by Zapier, Zapier MCP and the Zapier SDK all draw from one shared task pool, “with no separate task budgets by product.” Crucially, per-action task rates now vary by complexity — most actions cost 1 task, but AI steps charge by the AI model tier you select, Code by Zapier costs more if you extend the runtime, each successful Zapier MCP tool call counts as 2 tasks, and each Lead Router lead counts as 5 tasks. Paid Platform plans (Free, Professional, Team, Enterprise) all carry the same feature set per tier but are priced by a sliding task-volume tier ranging from 100 to 2,000,000 tasks per month; the headline price you see is the entry 100-task tier, and the monthly price rises as you pick a higher tier. The two AI products — Agents (metered by a separate activity quota, not tasks) and Chatbots (metered by number of chatbots) — remain separately priced add-ons layered on top of the Platform.

• Platform metering unit: tasks per month (100 / 2K / 10K / 100K / 1M / 2M selectable tiers); one shared pool across Zaps, AI steps, code, MCP & SDK
• Per-action task rate: 1 task for most actions; AI steps vary by model tier; Code by Zapier 1 task (more for extended runtime); MCP tool call = 2 tasks; Lead Router lead = 5 tasks
• Professional: from $19.99/mo billed annually ($29.99/mo monthly) at 100 tasks; unlimited multi-step Zaps and Premium apps
• Team: from $69/mo billed annually ($103.50/mo monthly) at 100 tasks; includes 25 users, SAML SSO
• Enterprise: contact-for-pricing; unlimited users, governance, VPC Peering, annual task pooling
• Agents add-on: Free (400 activities/mo) / Pro $33.33/mo annual ($400/yr; 1,500 activities/mo) / Enterprise contact — Agents draw on a separate activity quota, not tasks
• Chatbots add-on: Free (2 bots) / Pro $13.33/mo annual ($160/yr; 5 bots) / Advanced $66.67/mo annual ($800/yr; 20 bots) / Custom
• Annual discount: 33% vs monthly; pay-as-you-go overage available when you exceed a task tier, capped at 3× the plan’s task limit before runs are held
• Currency: 19 currencies selectable (USD, EUR, GBP, INR, JPY, and more)

What makes this different: Zapier prices the platform by consumption (tasks) rather than by seat — even the Team plan’s 25 users are bundled into a task-tier price. After the June 2026 unification, AI steps, code, MCP and SDK no longer have their own meters — they all consume from the same shared task pool at variable per-action rates — while Agents (activities) and Chatbots (chatbot count) stay on their own units, so a customer can still be metered on multiple dimensions at once.

Pricing by product

Zapier sells one usage-metered Platform and two AI add-ons. All Platform prices below are quoted at the entry 100-tasks-per-month tier; selecting a higher task tier raises the monthly price on the same plan. Annual billing (shown) saves 33% versus monthly. As of June 2026, the Platform task pool is shared across Zap workflows, AI steps, Code by Zapier, MCP and the SDK — there are no per-product task budgets — and per-action task rates vary (most actions 1 task; AI steps by model tier; MCP tool call = 2 tasks; Lead Router lead = 5 tasks).

Platform (Individual plans)

Platform (Business plans)

Agents (separately-metered AI add-on)

Chatbots (separately-metered AI add-on)

Sales motions across products: PLG / self-serve for Free, Professional, Team and the Agents/Chatbots Free–Advanced tiers; sales-led for Platform Enterprise, Agents Enterprise (coming soon), and Chatbots Custom.

Hidden costs : How tasks multiply and what they cost at volume

The $19.99 Professional headline is the 100-task entry tier, and it understates what an active automation account pays — because the meter ticks on every external-app action, not per workflow. Two patterns drive most of the surprise: multi-step Zaps that multiply task consumption, and how steeply that consumption costs versus cheaper alternatives once volume climbs.

Archetype 1 — a multi-step Zap multiplies tasks, and some steps now cost more than one. A task is any successful action that runs, so a single trigger that fans out to four actions burns at least four tasks per run. Since January 2024, built-in tools (Filter, Formatter, Paths, Delay, Storage, Tables) no longer count. But after the June 2026 meter unification, AI steps, code, MCP and SDK now consume from the same task pool at variable rates — an AI step’s cost depends on the model tier, a long-running code step costs more than 1 task, each MCP tool call is 2 tasks, and each Lead Router lead is 5 — so the same workflow can burn more tasks than its step count suggests:

Because price scales with the selected task tier, a couple of moderately wide Zaps push an account from the entry tier deep into the slider — and if you blow past the tier mid-month, overage bills on pay-per-task until you upgrade, capped at 3× your plan’s task limit, after which new Zap runs are held until the billing cycle resets.

Archetype 2 — the “convenience tax” at volume. Zapier’s per-task meter is the most expensive automation unit on the market once runs get into the tens of thousands. Community migration reports are consistent:

This is why “Zapier → n8n” is the dominant migration above ~50K tasks/month, and why Trustpilot and Reddit threads surface surprise $400–$1,200 charges after a misconfigured Zap looped overnight — the flip side of a meter that never stops automations when you exceed a tier.

Want to estimate your own Zapier bill? Use the Zapier pricing calculator to model monthly cost from task volume and plan tier — and see why AI cost unpredictability causes bill shock.

Pricing evolution : From a five-tier fixed-task lineup to an AI-orchestration slider

Zapier’s pricing held a stable five-tier, fixed-task shape through 2022–2023, then changed sharply in 2024: it refined what counts as a task, retired its entry Starter plan, and rebuilt the lineup around a task-volume slider before repositioning the whole platform around AI orchestration in 2025.

Cadence

Tracked range: 2021-03 – 2026-06. Both “adjustment” markers (Jan 2024, June 2026) flag meter redefinitions that changed what a task costs without changing headline plan prices. Dollar figures are taken from Wayback snapshots (2023) and the live 2026 capture; per-tier slider prices above the entry tier are not independently legible and are not guessed.

Notable changes

• 2021-03-08 — $5B valuation via a Sequoia-led secondary; bootstrapped on ~$1.4M, profitable since 2014. (Forbes.)
• 2022–2023 — Wayback shows a steady five-tier lineup with a fixed task bucket per plan (Free 100 / Starter 750 / Professional 2K / Team 2K / Company custom) and a “Save 33%” annual toggle.
• 2024-01 — Zapier reworked the meter: pay-per-task overage at 1.25×, unlimited Zaps on Free/Professional, and Filter, Formatter, Paths, Delay, Looping, Storage, Tables and Interfaces all stopped counting as tasks. (Zapier blog.)
• 2024-04-02 — The five-tier lineup collapsed: Starter was discontinued and upgraded to Professional at no cost, and the Company tier became a new Enterprise plan with VPC Peering, annual task pooling, and a Technical Account Manager.
• 2025 — Zapier layered Agents, Chatbots, and MCP onto the platform and reframed itself as an “AI orchestration” layer; the September 2025 Copilot launch put plain-English Zap building on every plan.
• 2026-06-02 — Zaps, Tables, Forms and MCP bundled into one plan; a task-volume slider (100 → 2M) re-prices Professional and Team; Agents and Chatbots remain separately-metered add-ons.
• 2026-06-24 / 2026-06-30 — Meter redefinition: AI steps, Code by Zapier, MCP and the SDK fold onto one shared task pool (“no separate task budgets by product”), and a task becomes “any successful action that runs” — but at variable per-action rates (AI steps by model tier, Code more for extended runtime, MCP tool call = 2 tasks, Lead Router lead = 5 tasks). Plan prices and task allowances were unchanged; pay-per-task overage now caps at 3× the plan’s task limit before runs are held. Agents stay on a separate activity quota. (Zapier Help — task usage rates.)

The 2024 consolidation in detail

The April 2024 change is the most consequential moment in Zapier’s pricing history. Before it, the lineup carried five tiers and a real entry rung — Starter at $19.99 for 750 tasks sat below a $49 Professional. Zapier deleted Starter outright, upgraded those customers to Professional for free, and made Professional the new $19.99 floor, simultaneously folding the legacy Company tier into a purpose-built Enterprise plan. Paired with the January 2024 meter refinements — overage billing, unlimited Zaps, and exempting built-in tools from task counting — the effect was to simplify the buying decision down to “how many tasks do you need,” which the later task-volume slider made literal. It also nudged the average paid customer up: the cheapest paid plan went from a $19.99/750-task Starter to a $19.99/100-task Professional, so the entry price stayed flat while the included volume at that price fell.

The June 2026 meter unification in detail

The June 2026 change rhymes with 2024: again, headline plan prices and task allowances held perfectly steady, and again what changed was the meter underneath them. Two things happened at once. First, a simplification: AI steps, Code by Zapier, MCP and the SDK lost their separate budgets and now draw from one shared, account-wide task pool — the pricing page’s promise that your pricing “works the same way everywhere you use Zapier.” That collapses what used to be several product-specific meters into a single number the buyer already understands.

Second, and quieter, a meter broadening. A task was previously framed as one successful external-app action; it is now “any successful action that runs in Zapier,” and the per-action rate is no longer flat. AI steps charge by the selected model tier, a long-running Code step costs more than one task, each MCP tool call counts as 2 tasks, and each Lead Router lead counts as 5. Because more kinds of work now draw from the same fixed task allowance — and some of that work draws at a multiple — the effective price per unit of value rises for AI-, code- and MCP-heavy accounts even though no number on the pricing page moved. An account that leaned on AI steps or MCP under the old separate-budget regime will reach its task limit sooner under the shared pool, and the overage cap moving to 3× the plan limit (from the prior 1.25×-rate framing) sets a higher ceiling before runs are held. This is the same playbook as the 2024 floor change: hold the sticker, move the meter.

What’s unique : The original per-task meter, now a shared pool with weighted rates

1. Priced per task — now “any successful action that runs” — not per workflow or seat. Zapier popularized the per-task meter for automation, and as of June 2026 it broadened the definition: a task is no longer just one successful external-app action but any successful action that runs, with weighted per-action rates (AI steps by model tier, MCP tool call = 2 tasks, Lead Router lead = 5). It is the granular opposite of n8n’s per-execution model, where a whole workflow run counts once regardless of step count, and a key reason Zapier is the value-metric reference point its rivals position against — but the weighting moves it closer to a complexity-priced meter than the flat 1-task-per-action unit it taught the category.

2. One shared task pool for the Platform, but still three meters across products. As of June 2026 the Platform’s many sub-meters collapsed into a single shared task pool — Zaps, AI steps, code, MCP and SDK all draw from the same allocation “with no separate task budgets by product.” Yet Agents (metered by “activities”) and Chatbots (metered by chatbot count) stay on their own units, so a single customer can still be billed on three different usage dimensions at once. The move simplifies within the Platform while leaving the cross-product multi-meter intact.

3. A task-volume slider decouples price from headcount. Paid plans are priced by a 100 → 2M task selector, not per seat; even the Team plan’s 25 users are bundled into the task-tier price. The bill scales with automation volume, not with how many people log in.

4. The meter cuts both ways — built-in tools exempt, but AI/MCP/code weighted up. Since January 2024, Filter, Formatter, Paths, Delay, Storage, Tables and Interfaces don’t count as tasks — internal plumbing is free. But the June 2026 redefinition pulled in the opposite direction at the high end: AI steps, code, MCP (2 tasks/call) and the SDK now consume from the same pool at weighted rates. Zapier still exempts the cheap internal work, but it now charges more for the expensive, compute-heavy actions — a meter tuned to track its own underlying cost, not a flat per-action unit.

5. Breadth and execution infrastructure as the moat, not price. With 9,000+ app integrations and 13 years of production-grade execution (retries, token refresh, credential storage), Zapier sells reliability and coverage rather than the cheapest unit — explicitly conceding the low-cost ground to Make and self-hosted n8n while defending the AI-agent orchestration layer on top.

Strengths & weaknesses

Billing UX : Task-tier slider, currency picker, and pay-as-you-go controls

• Task-tier slider — an interactive “How many tasks do you need?” control with notches at 100, 2K, 10K, 100K, 1M, and 2M that re-prices the Platform plans live as you drag.
• Shared task pool — a single account-wide task allocation; the pricing page states “Zap workflows, AI steps, code, MCP, and SDK all draw from the same task allocation, with no separate task budgets by product.”
• Variable per-action task rates — “every step in a Zap and every external connector call uses tasks — though rates may vary”; AI model tier, code runtime, and connector type each affect how many tasks an action uses (AI by model tier, MCP = 2 tasks, Lead Router = 5 tasks).
• Pay monthly / Pay yearly toggle — switches all displayed plan prices; the yearly state is labeled “Save 33%” and paid tiers show “Billed annually.”
• Pay-as-you-go overage with email alerts — exceed your tier and you’re switched to pay-per-task (unless you turn it off); Zapier emails at 80% and 100% of the limit, which caps at 3× the plan’s task limit before runs are held until the cycle resets.
• Usage alerts & CSV export (Team/Enterprise) — admins can export a per-member task-usage CSV, and Enterprise can set usage alerts at a defined threshold.
• Currency selector — a 19-currency dropdown (USD, AUD, BRL, CAD, CHF, DKK, EUR, GBP, HKD, ILS, INR, JPY, MXN, NOK, NZD, PLN, SEK, SGD, ZAR) re-prices plans in local currency.
• Annual task limits (Enterprise) — Enterprise pools tasks annually instead of letting them expire monthly, so unused monthly task budget carries forward across the year.
• Per-add-on billing toggles — the Agents and Chatbots add-on tabs each carry their own independent Pay-monthly/Pay-yearly toggle and currency picker, billed separately from the Platform plan.

Strategic wins : Why the per-task meter and bootstrapped model worked

1. Inventing the per-task meter as the category’s value metric

Zapier made “tasks” the default way to price automation, and the unit is intuitive for the buyer it courts: one action moved between apps, one task. That legibility helped it reach $100M ARR in under a decade and become the value-metric reference everyone else benchmarks against. The meter’s weakness at high volume is real, but its clarity at low-to-moderate volume is exactly what won the self-serve market.

2. Pricing for profitability on a bootstrapped balance sheet

Zapier reached a $5B valuation having raised only ~$1.4M and being profitable since 2014 — a discipline that shows up in the pricing. Without VC pressure to chase logo growth at any cost, Zapier could price the meter to fund operations and let usage-based expansion compound, the land-and-expand logic of usage-based SaaS executed without burning capital.

3. Refining the meter to charge only for value-creating actions

The January 2024 decision to stop counting Filter, Formatter, Paths and other built-ins as tasks gave customers back free internal plumbing and concentrated the bill on external-app actions. Charging for the work that creates value — and visibly removing charges for the work that doesn’t — is a trust move that the better usage-based pricing playbooks recommend, and it softened a long-standing complaint that “everything counts.”

4. Repositioning to AI orchestration to defend a premium meter

By reframing from “automation” to “AI orchestration” — Agents, Chatbots, MCP, and a plain-English Copilot — Zapier moved the value story to governance, breadth, and reliability rather than cost-per-task, the ground it cedes to Make and self-hosted n8n. It is a deliberate bet that being the safe, broad orchestration layer for AI-agent workflows justifies a premium meter even as cheaper rivals undercut it.

5. Folding AI, code, MCP and SDK onto one shared task pool — and weighting it to cost

The June 2026 unification collapsed several product-specific meters into a single shared task pool — “your pricing works the same way everywhere you use Zapier” — so a buyer reasons about one number instead of separate AI-step, code and MCP budgets. The clever part is doing it while holding the sticker price and task allowances flat: the change reads as a simplification, yet the simultaneous shift to weighted per-action rates (AI by model tier, MCP ×2, Lead Router ×5) quietly re-couples the meter to Zapier’s own compute cost on the value-metric front. It is a textbook way to reprice the expensive actions without touching a published number — though it leans on buyers not modelling the effective hit.

Areas to improve : Transparency, bill shock, and the volume cliff

1. Publish per-tier slider prices

Only the entry 100-task price ($19.99 Professional, $69 Team) is shown up front; the cost at 10K, 100K, or 1M tasks appears only as you drag the slider. The June 2026 weighted rates make this worse: with AI steps charged by model tier and Code by Zapier “more for extended runtime,” a buyer can no longer convert an action count into a task count without a published rate card per model and runtime band. Publishing the full per-tier price ladder and the per-action task-rate table — not just “rates may vary” — would let buyers model their real effective cost before committing, the transparency the usage-invoicing playbooks call for.

2. Harden the guardrails against bill shock

Pay-as-you-go overage keeps automations running, but it is also how a looping or misconfigured Zap turns into a $400–$1,200 surprise — and the June 2026 shared pool widens the blast radius, since AI steps (by model tier), MCP (2 tasks/call) and Lead Router (5 tasks/lead) now drain the same budget at multiples. Capping pay-per-task at 3× the plan limit is a real guardrail, but 3× of a high task tier is still a large number; surfacing dollar spend caps, approaching-limit alerts, and an anomaly circuit-breaker on the weighted actions would prevent the runaway-bill stories that drive churn — exactly the thresholding and alerting controls usage-priced vendors are expected to ship.

3. Address the volume cliff against Make and n8n

At tens of thousands of tasks a month, Zapier is 4–15× pricier than Make and far more than self-hosted n8n, and migration off Zapier is the standard response. A finer overage curve, a high-volume committed-use band, or an execution-style “bundle the multi-step Zap” option would slow the bleed of high-volume accounts without abandoning the per-task model that wins everyone else.

Key takeaways

1. Your value metric becomes the category’s benchmark — for better and worse. Zapier’s per-task meter is intuitive enough to define how automation gets priced, but granular enough that every cheaper rival positions against it. The unit that wins the entry market can become the liability at scale.
2. Removing charges can build more trust than adding features. Exempting built-in tools from task counting in 2024 gave customers back free plumbing and directly answered the “everything counts” complaint — a subtractive change that improved perceived fairness.
3. A flat sticker price can hide a quieter increase — twice. The 2024 consolidation kept the cheapest paid plan at $19.99 but cut its included tasks from 750 to 100; the June 2026 unification held every price and allowance steady yet broadened what a task is and weighted AI/MCP/code actions above 1× — both times the effective price-per-unit rose while the headline stayed put. Repricing the meter instead of the number is Zapier’s recurring move.
4. Overage that never stops the work is a double-edged default. Pay-as-you-go protects uptime but is also the mechanism behind surprise bills, and a shared pool that AI/MCP/code drain at multiples reaches the limit faster; predictable usage pricing needs guardrails — the 3× cap is one — as much as continuity.
5. Bootstrapped discipline shows up in the meter. Profitable since 2014 on ~$1.4M raised, Zapier could price for margin and expansion rather than land-grab growth — a reminder that capital structure shapes pricing strategy.

UBP implications

1. Granular meters win the entry market and lose the volume market. A per-action unit is easy to understand and cheap to start, but it compounds fast — so the same legibility that drives adoption drives high-volume churn unless the curve is tamed.
2. You can reprice a usage meter without touching a published price. Zapier broadened its task definition and weighted AI/MCP/code actions above 1× in June 2026 while holding every plan price and task allowance flat — a packaging/metering change, not a price change. Redefining the billable unit and its rate is a quieter lever than a sticker hike, and one regulators and procurement teams scrutinise far less, but it shifts effective price all the same.
3. A shared pool simplifies the buyer’s mental model but raises consumption — and needs spend controls. Collapsing several product meters into one task budget makes pricing legible, yet every product now competes for the same allowance, so customers reach their limit sooner; auto-overage then keeps automations alive but creates bill-shock risk. Pair the simplification with caps, alerts, and anomaly detection rather than treating them as optional.

Sources

• Zapier pricing page (accessed 2026-06-30)
• Zapier Enterprise (accessed 2026-06-30)
• Zapier Help — How is task usage measured in Zapier? (accessed 2026-06-30)
• Zapier blog — 2024 plan improvements (accessed 2026-06-02)
• Zapier Help — Enterprise plan and existing plan updates (accessed 2026-06-02)
• Zapier Help — How to select your Zapier plan (accessed 2026-06-24)
• Forbes — Zapier bootstraps to a $5B valuation (2021-03-08) (accessed 2026-06-02)
• Sacra — Zapier revenue, valuation & funding (accessed 2026-06-02)
• Wayback Machine pricing snapshots (zapier.com/pricing, 2022-01 – 2023-12) — see the pricing-evolution timeline above
• Compare other automation pricing in the pricing blueprint

Bottom line

Zapier prices automation on the meter it popularized — tasks — across a Free tier, a task-volume slider for Professional and Team, sales-led Enterprise, and two separately-metered AI add-ons (Agents and Chatbots). As of June 2026 a task is “any successful action that runs,” and AI steps, code, MCP and the SDK all draw from one shared pool at weighted rates (AI by model tier, MCP ×2, Lead Router ×5) — a simplification that holds every sticker price flat while quietly raising the effective price for AI-heavy accounts. The per-task unit is the clearest on-ramp in the category and its biggest liability at scale: intuitive and cheap to start, 4–15× pricier than Make and far more than self-hosted n8n once volume climbs.

Want to compare Zapier against other automation and AI pricing models? Browse the pricing blueprint.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

Unified task meter: AI steps, code, MCP & SDK now share one task pool (current)

Jun 2026

Zapier moved AI steps, Code by Zapier, Zapier MCP and the Zapier SDK onto the same task-based meter — 'AI steps, code, and SDK now all follow the same task-based pricing model.' Tasks are now a single shared account pool ('no separate task budgets by product'), and per-action task rates vary by complexity: a task is 'any successful action that runs,' but AI steps charge by selected model tier, Code by Zapier charges more for extended runtime, each Zapier MCP tool call counts as 2 tasks, and each Lead Router lead counts as 5 tasks. Plan prices and task allowances were unchanged (Free $0/100 tasks, Professional $19.99 annual, Team $69 annual, Enterprise custom); what a 'task' costs changed. Pay-per-task overage now caps at 3× the plan's task limit before runs are held. Agents remain on a separate activity quota.

Unified 'AI orchestration' plan on a task-volume slider (current)

Jun 2026

Zaps, Tables, Forms and Zapier MCP are bundled into Free, Professional and Team at no extra cost. A single task-volume slider (100 → 2M tasks/mo) re-prices the paid plans: Free ($0, 100 tasks), Professional (from $19.99/mo annual; $29.99 monthly), Team (from $69/mo annual, 25 users; $103.50 monthly), Enterprise (custom). Agents and Chatbots remain separately-metered add-ons. Annual billing saves 33%.

AI-orchestration repositioning — Agents, Chatbots, MCP, Copilot

Sep 2025

Zapier reframed from 'automation' to an 'AI orchestration platform,' layering two separately-metered AI add-ons (Agents, metered by activities; Chatbots, metered by chatbot count) plus Zapier MCP (an AI action layer over 9,000+ apps) on top of the task-metered platform. A September 2025 Copilot launch let users build Zaps from plain-English prompts on every plan, including Free.

Starter discontinued; Professional becomes entry tier; Company → Enterprise

Apr 2024

Zapier collapsed its five-tier lineup: the Starter plan was retired and those users upgraded to Professional at no extra cost, making Professional ($19.99/mo annual) the entry paid plan; the old Company tier was replaced by a new Enterprise plan with advanced admin controls, VPC Peering, annual task pooling, and a Technical Account Manager. Reported ARR was ~$310M. (Zapier blog, 2024-04-02.)

Task-counting overhaul + pay-per-task overage

Jan 2024

Zapier made Zaps unlimited on Free and Professional, introduced pay-as-you-go overage at 1.25× the subscription rate when you exceed your task tier, and stopped counting built-in tools as tasks — first Filter, Formatter and Paths (Jan 8), then Delay, Looping, Sub-Zap, Digest, Zapier Manager, Storage, Tables and Interfaces (Jan 18). Only actions that touch an external app now consume a task.

Five-tier fixed-task lineup (Free/Starter/Professional/Team/Company)

Jan 2023

Wayback snapshots through 2022–2023 show a five-tier structure with a fixed task bucket per plan: Free ($0, 100 tasks), Starter ($19.99/mo annual, 750 tasks), Professional ($49/mo annual, 2K tasks), Team ($69/mo annual, 2K tasks, unlimited users), and Company (contact sales). Annual billing already advertised 'Save 33%.'

$5B valuation on ~$1.4M raised — bootstrapped and profitable

Mar 2021

A Sequoia-led secondary valued Zapier at $5B in early 2021 on roughly $140M ARR, despite the company having raised only ~$1.4M in venture funding and being profitable since 2014. No change to the task-metered pricing model; the milestone underwrote Zapier's bootstrapped, founder-controlled trajectory. (Forbes, 2021-03-08.)