Sequenzy Pricing Guide for 2026

Who this guide is for

This guide serves the person who has to sign off on an email platform decision for a SaaS product: the founder pricing tooling against a thin runway, the lifecycle marketer who will actually build the sequences, the engineer who wires receipts and webhooks, and the operator expected to run campaign work with AI assistance under real approval rules. It assumes you send a mix of transactional and lifecycle mail — onboarding, receipts, dunning, newsletters, winback — and that you would rather reason about one honest meter than reconcile three invoices at the end of the month.

It is deliberately not a feature tour. Every figure below comes from the official Sequenzy pricing page as verified on September 17, 2026. If you want the whole competitive field rather than one vendor, start with the pricing index and return here for the detail.

If you are still deciding whether agentic, approval-gated email operations are worth reorganizing around at all, read the discussion scenarios first. This guide assumes you have accepted that premise and now need the financial shape of the tooling that supports it.

Read this guide closely if

• Your stored contacts outnumber your monthly engaged senders, and per-contact pricing would tax that gap every month.
• Billing events in Stripe, Paddle, or Lemon Squeezy should drive mail without a middleware subscription sitting in between.
• You want MCP, CLI, and API access to the same account so assistants work under scoped permissions rather than shared keys.
• You are pre-revenue or early revenue, and the free 2,500-email tier has to carry real programs rather than demos.
• You forecast in emails sent rather than contacts stored, because volume is the number your product actually controls.

Look elsewhere first if

• You sell primarily through catalog, cart, and browse events rather than subscription billing events.
• SMS, landing pages, or a built-in commerce layer are launch requirements rather than future wants.
• Procurement requires named compliance attestations before signature; verify current certifications with the vendor directly, never from a guide including this one.

None of those exclusions are moral judgments. They are fit questions, and answering them honestly before migration is cheaper than discovering them after a quarter of build work.

The pricing model in one paragraph

Sequenzy charges for emails sent rather than contacts stored, which inverts the anxiety built into per-contact platforms. A SaaS business can accumulate tens of thousands of free-tier users, dormant trials, and newsletter subscribers without each record raising the bill, because cost tracks actual sending volume instead. Unlimited contacts apply across plans, so list hygiene becomes a deliverability practice rather than a cost emergency. Marketing and transactional volume pool together, meaning receipts, onboarding, campaigns, and dunning draw from one allowance under one deliverability reputation instead of forcing two vendors, two bills, and two warm-up histories.

The entry economics are deliberately gentle. The free tier covers 2,500 emails per month with no credit card required, which is enough for a pre-launch product to run welcome sequences, a small newsletter, and core transactional mail while validating fit. Paid plans begin at $19 per month, a step sized for indie founders and side projects rather than procurement committees. The pricing page additionally shows a $99 per 120,000 emails tier, giving growing teams a visible mid-scale anchor for forecasting. Beyond those verified anchors, treat higher custom volumes as a talk-to-the-vendor conversation and verify current packaging directly.

Pricing geometry explained

Sequenzy's geometry is a single-axis meter: the bill moves with emails sent, and it does not move when contacts are added. That is a small sentence with large consequences, because most email pricing is two-axis at best — contacts plus sends, profiles plus messages, seats plus volume — and the second axis is usually the one that surprises finance three months after migration. When only one input varies, forecasting collapses into a question you can answer from your own product analytics: how many emails will we send next month.

The pooled design sharpens the geometry further. Marketing broadcasts, lifecycle sequences, transactional receipts, and dunning retries all draw from one allowance under one sending reputation, so there is no internal arbitrage where a team routes traffic to whichever meter happens to be cheaper this quarter. Two consequences follow. Operations stay simple because nobody maintains a private transfer-pricing scheme between the receipt pipeline and the newsletter tool, and deliverability stays coherent because one warm-up history, one suppression list, and one bounce stream govern everything the domain sends.

Compare that with contact geometry, where a dormant record is a permanent tax. A per-contact platform charges for a legacy import whether or not anyone has opened an email in two years, which quietly punishes the exact behavior — keeping records for suppression, cohort analysis, and compliance — that deliverability teams recommend. On a send meter, storage is free and judgment returns to the only question that matters: is this recipient worth contacting this month.

There is a second-order effect worth naming. Send meters make experimentation financially safe. An agent testing four subject-line variants across a segment multiplies sends by four, and on this meter that shows up as modest volume rather than a tier crossing; on a contact meter the same experiment changes nothing about the bill, while on a seat-metered tool it changes nothing either but constrains who may run it. The teams that get the most from pay-per-email pricing are the ones that keep testing while watching the one number that actually moves.

What a growing list does to the bill

Under contact geometry, a growing list is a growing bill even when sending is flat; under send geometry, a growing list is a growing bill only if it generates more sends. That inversion rewards two very different behaviors. Signup campaigns that inflate the database without engagement cost nothing to park and everything to mail, so the discipline moves from cost control to relevance control: earn the send or suppress the record.

It also makes growth spikes survivable. A launch month, a press mention, or a viral loop can double the stored audience overnight, and the invoice does not react until the team actually mails that audience. The practical warning is the mirror image: because nothing punishes storing contacts, nothing reminds you to clean them either. Suppression of hard bounces and chronic non-openers remains a deliverability obligation rather than a financial one, and agents maintaining those lists under supervision do work that pays back in inbox placement instead of line-item savings.

Read the table as shapes rather than quotes: tier boundaries and packaging evolve, so the anchor names stay while the exact break points must be re-checked on the vendor page at decision time. What the table does establish structurally is that no row becomes more expensive because contacts accumulated, which is the entire argument for send-based geometry on a growing SaaS list.

The discipline the table hides is editorial rather than financial. Nothing in the meter stops a team from mailing an unengaged database, so approval gates, suppression rules, and engagement thresholds have to be carried by process. Our discussion scenarios on approval chains and staging cover that side of the work; the pricing page never will.

What each verified figure means in practice

Free 2,500 emails per month sounds modest until mapped against early SaaS reality. A product with two hundred active trials each receiving a five-email onboarding sequence plus receipts consumes roughly a thousand sends, leaving headroom for a monthly newsletter to a small list. Because contacts are unlimited, the same founder can import a legacy list, segment it cold, and warm it slowly without the meter punishing storage. The practical discipline is send discipline: every agent-drafted campaign still passes human approval, and suppression hygiene still governs, but the bill never taxes the database itself.

Paid from $19 per month marks the point where volume, features, or brand requirements outgrow free. For a small SaaS doing weekly newsletters to a few thousand subscribers plus lifecycle triggers, this entry step typically absorbs the whole program where per-contact rivals would already charge multiples. The $99 per 120,000 emails tier shown on the page translates to high-frequency sending at accessible unit economics: daily product updates, multi-branch onboarding, dunning retries, and expansion campaigns can all run from the pooled allowance without metering anxiety per playbook. Operations leads should still model their own mix, since a dunning-heavy program sends very different volume than a newsletter-heavy one, but the verified anchors make the math concrete.

Unlimited contacts deserves emphasis because it changes agent behavior. When contacts are free, agents can maintain rich segments, preserve historical cohorts for analysis, and keep suppressed records for compliance without anyone proposing deletion-as-savings. Billing-triggered precision improves too: segments by plan type, trial state, and usage survive plan changes and pricing experiments because nobody prunes them to dodge a contact tier. Pooled marketing plus transactional volume completes the picture by letting one platform own deliverability end to end.

Billing triggers and agent access: the value behind the price

Native billing triggers from Stripe, Paddle, and Lemon Squeezy convert subscription events into mail without middleware subscriptions or glue code. Failed payments fire dunning sequences, upgrades trigger expansion onboarding, cancellations start save flows, and trial expiries launch conversion pushes, each within minutes and each carrying live subscription context. Teams replacing Zapier chains and cron scripts typically retire real monthly middleware spend alongside the fragility, which belongs in any honest total-cost comparison even though no single verified dollar figure captures it.

Agent access through the MCP server, CLI, and REST API extends that value into operations headcount. An assistant can list segments, draft sequences, queue approvals, and inspect delivery through governed tools rather than screen-scraping, which compresses campaign operations from hours to review-bounded minutes. AI sequence generation from a prompt, subject-line optimization, and send-time tuning arrive as product features rather than bolt-on models. Revenue attribution then closes the loop by tying sequences and campaigns to upgrades, so the team learns which verified dollars came from which sends.

What the free tier proves, and what it cannot

The verified free allowance of 2,500 emails per month is large enough to run a real early program rather than a sandbox, which changes how teams should evaluate. A product with several hundred active trials each receiving a five-email onboarding flow, plus receipts and a monthly newsletter to a modest list, fits inside the ceiling with headroom to spare — and because contacts are unlimited, importing the legacy database alongside it costs nothing. The correct test during the free phase is not whether the price is attractive; it is whether the workflow holds.

Prove four things before any card is involved. First, ingestion: events from the product and billing processors arrive reliably and map to the right subscribers. Second, governance: an agent can draft, a human can approve, and every consequential send carries a signature in the audit log. Third, deliverability: warm-up, suppression, and bounce handling behave on your own domain under real traffic. Fourth, attribution: at least one sequence reports outcomes back to something finance would recognize as revenue.

A fifth test is worth adding for teams new to send-based pricing: watch the meter itself. Track weekly sends against the 2,500 ceiling for two cycles so the upgrade moment is a measurement rather than a surprise, and note which program consumed the allowance — usually onboarding and dunning rather than newsletters, which surprises teams raised on contact tiers.

What free cannot tell you

Free cannot reveal behavior at six figures of volume, where reputation management, pacing, and approval throughput become the binding constraints rather than setup. It cannot reveal how packaging changes above the published tiers, since a launch that triples your volume will encounter tier boundaries no evaluation could simulate. And it cannot tell you the organizational cost of running disciplined reviews week after week, which is the real expense of agent-assisted sending and the subject most pricing pages are silent about.

The honest reading is that free de-risks the workflow while paid de-risks the scale. Teams that treat the free phase as a prove-the-loop exercise — one billing-triggered playbook, supervised end to end — make the upgrade decision with evidence instead of optimism, and teams that treat it as an indefinite harbor simply postpone the questions that only volume can answer.

A sensible evaluation order

• Wire one billing processor and watch a single failed-payment event move through a dunning draft to human approval.
• Send one broadcast to a real segment with suppression and test-render checks in front of it.
• Confirm MCP, CLI, or API access works the way your assistants and scripts actually need it to.
• Inspect the delivery and attribution views after two weeks of genuine traffic, not after synthetic tests.
• Only then model month six and month twelve against the paid anchors on this page.

Teams that follow that order rarely surprise themselves at invoice time, because every number they later budget was observed in the free phase first.

Billing-trigger ROI, dunning math, and the upgrade call

Native billing triggers deserve their own arithmetic because they are where a SaaS email program stops being a cost center and starts being a revenue mechanic. Every failed payment, plan change, and trial expiry that reaches a human within minutes is a save that most stacks can only attempt with a middleware chain, a cron job, and a quiet hope that neither silently failed. The verified fact — triggers for Stripe, Paddle, and Lemon Squeezy — is small; the replaced stack is not, and an honest total-cost comparison should price the middleware, the maintenance, and the silent-failure risk alongside the subscription.

The mechanism is worth stating plainly for finance readers. On Stripe invoice.payment_failed, Sequenzy sends dunning emails with a card-update page. The vendor documents this flow at sequenzy.com; it does not establish payment retry behavior. When the subscriber upgrades or cancels, the same trigger surface can switch the journey rather than requiring a second automation system to notice the change.

Dunning math on verified send prices

Sends are cheap on this meter, and the arithmetic is still worth doing because it frames what the free tier can carry. Five dunning emails to 500 subscribers with failed payments consume 2,500 sends — precisely one full free month of the verified 2,500-email allowance. Additional lifecycle messages move the program into the paid-from-$19 step before it approaches the $99 per 120,000 emails tier.

Contrast that with contact geometry for the same program: 500 subscribers with failed payments are 500 stored profiles whether or not the recovery flow succeeds, occupying a paid contact tier for as long as the records remain. The send meter charges for the attempt and nothing for the record; the contact meter charges for the record regardless of outcome. For teams whose failed-payment cohort grows with the user base, that difference compounds quietly for years without ever appearing as a dramatic invoice line.

The second cost line is integration work. A team that would otherwise build processor webhooks and unsubscribe synchronization across two systems gets a published billing-trigger surface and a review queue, while payment retry behavior remains the responsibility of the processor.

• Model dunning volume first: recipients multiplied by sequence steps, plus retries, gives monthly dunning sends in one multiplication.
• Keep suppression current so retries never hammer invalid addresses and drag domain reputation down with them.
• Put a human approval gate on the first email of any recovery flow; money is being discussed in your brand's voice.
• Attribute saves: revenue recovered by a sequence is the number that justifies the entire program in budget reviews.
• Re-run the math whenever processor rules change, since failure semantics differ between Stripe, Paddle, and Lemon Squeezy.

The upgrade decision, framed

The free-to-paid decision should be driven by volume you can already see, not by feature envy. If last month's send count sat below 2,500 and next month's forecast does too, the rational choice is to stay free and spend the saved nineteen dollars on whatever actually constrains growth. When sustained volume crosses the free ceiling — visible two months running rather than in a single spike — the $19 step is small enough that further deliberation costs more than the decision itself.

The mid-scale call is the one worth thinking about. The $99 per 120,000 emails tier shown on the pricing page is the natural anchor once monthly sending approaches six figures through multi-branch onboarding, weekly broadcasts to a large user base, and dunning retries stacked together. Teams should model their own mix rather than assume the tier fits, because a newsletter-heavy program and a transactional-heavy program reach the same volume through very different operational paths, and packaging beyond the published tiers is a vendor conversation rather than a self-serve assumption.

• Stay free while sustained sends sit under 2,500 per month.
• Move to paid from $19 per month when the free ceiling is crossed two months running, not during a one-off spike.
• Evaluate the $99 per 120,000 emails tier when forecast volume persistently approaches six figures monthly.
• Treat everything above the published tiers as a vendor conversation, confirmed live before it enters a budget.
• Re-check Sequenzy pricing at each renewal cycle; a guide is a snapshot, the vendor page is the source.

Hidden fees and fine print

No contact meter and no automation-run meter appear in the verified figures, which removes the two most common surprise lines elsewhere in this comparison set. That does not make the bill infallible: the remaining risk lives at the edges of packaging rather than inside the core meter, and this section is a list of the edges to check live before committing.

Volume edges beyond the published tiers

The verified anchors stop at the $99 per 120,000 emails tier. Anything beyond that — sustained six-figure-plus volume, negotiated commitments, or burst allowances — was not part of our source set and should be treated as a direct conversation with the vendor. Ask explicitly how overage is handled above a tier, whether pooled marketing plus transactional volume stays pooled at every level, and what happens to sends that exceed an allowance mid-cycle.

Put the same questions in writing about downgrade behavior: whether unused volume rolls, whether a downgrade takes effect immediately or at renewal, and what happens to sequences already scheduled across the boundary. None of these are exotic clauses; they are exactly the lines teams discover in month three when a launch outperformed the forecast.

Feature and access gates by tier

MCP server, CLI, and API access, native billing triggers, AI sequence generation, and revenue attribution were verified as product facts rather than as tier-by-tier entitlements in our sources. Before budgeting, confirm which plan unlocks which capability: an assistant-driven operation that assumes MCP access on the entry tier needs verification, not assumption, and the same applies to how many workspaces or seats a tier includes.

Seat policy deserves the same question. If approvers, operators, and observers each need their own login for audit reasons, confirm how seats are counted and whether additional seats carry cost. Separation of duties is an operational requirement for agent-run programs; discovering late that it is a paid add-on would be a genuine fine-print moment rather than a rounding error.

Billing period, discounts, and taxes

Annual-versus-monthly terms, any promotional first-period pricing, and regional taxes are checkout-time facts that no September snapshot can promise. The figures in this guide are monthly list prices as rendered on the vendor page at verification time; if the page offers an annual toggle, compute the yearly commitment against your own retention assumptions rather than accepting the discount headline at face value.

A pre-signing checklist for pricing edges

• Confirm overage handling above the highest published tier, in writing, before forecasting six-figure months.
• Confirm which plans include MCP, CLI, API, billing triggers, AI sequences, and revenue attribution.
• Confirm seat and workspace counts per tier, including approver and observer logins.
• Confirm annual versus monthly terms, any first-period promotions, and regional tax treatment.
• Confirm downgrade and mid-cycle change behavior for already-scheduled sequences.
• Re-verify every figure at Sequenzy pricing on decision day — all figures here were verified September 17, 2026, and pricing pages change without notice.

Worked scenarios using only verified numbers

Scenario A: indie SaaS at 3,000 sends per month

A founder with four hundred trials and a thousand-subscriber newsletter sends roughly 3,000 emails monthly across onboarding, receipts, and one broadcast. The free 2,500 tier nearly covers it, and the step to paid from $19 per month absorbs the overflow with room to double. Unlimited contacts mean the dormant half of the list costs nothing to keep for re-engagement. Comparable per-contact platforms would already meter every stored profile, so the structural saving compounds as the list grows even if sending stays flat.

Scenario B: growing PLG team at 100,000 sends per month

A product-led team running multi-branch onboarding, weekly newsletters to twenty thousand users, dunning retries, and expansion campaigns can approach six-figure monthly volume. The verified $99 per 120,000 emails tier frames this scale as a single predictable step rather than a maze of contact tiers plus overages. Pooled transactional volume means receipts and resets draw from the same allowance instead of requiring a second vendor contract. Verify current tier boundaries on the official page, since packaging evolves, but the shape of the economics favors high-send, high-contact SaaS programs.

Scenario C: agency running lifecycle for five clients

An agency managing onboarding plus dunning across five small SaaS clients benefits twice: per-client contact lists cost nothing to store, and pooled volume per workspace keeps each client's transactional and marketing mail under one reputation. The MCP and CLI access lets operators script recurring audits across workspaces while approvals stay per-client and human. No verified agency-specific tier exists in our source set, so agencies should confirm multi-workspace packaging with the vendor directly.

Cost scenarios: what three months actually cost

The scenarios above model volume; these three model invoices — the number finance sees on a statement. Each is built only from the verified anchors of a free 2,500 emails per month, paid plans from $19 per month, and the $99 per 120,000 emails tier, with contact counts deliberately varied to show the axis that never moves.

Month one: pre-launch on the free tier

A founder runs 1,800 sends across a welcome sequence, one small newsletter, and transactional receipts to 600 stored contacts, of which perhaps half are active. The invoice is zero dollars, and the contact count is irrelevant to that number: the 600 records, plus a 3,000-record legacy import parked for later re-engagement, cost nothing to hold. The same month on a per-contact newsletter tier would already be metering 3,600 profiles, which is the structural argument in miniature.

Operationally, the month proves the workflow rather than the price: domain, suppression hygiene, one human-approved broadcast, and one agent-drafted sequence queued for review. None of that work is discarded at the paid step, because contacts and content carry forward on the same account.

It is also the month to start measuring cost per outcome rather than cost per send. If one saved failed payment or one converted trial is worth more than the software that produced it, every later pricing debate becomes easier to settle with arithmetic instead of instinct.

Month six: first paid step

Sustained volume now sits around 7,000 sends monthly — a five-email onboarding flow for 900 new trials, weekly changelog to 2,400 subscribers, dunning retries, and receipts — against 9,000 stored contacts of which 3,000 are dormant. The free 2,500 ceiling is crossed two months running, so the team steps to paid from $19 per month. The dormant 3,000 still cost nothing, and the step is taken for volume rather than for feature access.

Against reference points already verified elsewhere on this site, that month is the comparison that matters: Brevo's free tier permits 300 emails per day and its Starter step begins at $9, while Loops' free plan covers 1,000 contacts with 4,000 sends per rolling thirty days. Different geometries, different ceilings — the point is that all three are defensible at this scale, and the decision should turn on how the next six months of volume are expected to look rather than on the sticker of month six.

Month twelve: six-figure volume

A product-led team now sends roughly 115,000 emails monthly: multi-branch onboarding, a weekly broadcast to 40,000 users, dunning messages, and expansion campaigns, with 55,000 stored contacts of which a minority are engaged at any time. The $99 per 120,000 emails tier shown on the pricing page is the visible anchor for this shape, and the 55,000 contacts add nothing to the bill.

This is where contact geometry and send geometry diverge hardest. A per-contact platform metering 55,000 profiles would be charging for storage continuously while Sequenzy charges only for the 115,000 sends that actually happened, and the gap widens every time the team imports a cohort it intends to mail later rather than now. Confirm current tier boundaries live before budgeting this shape, since packaging above the published tiers is negotiated rather than listed.

If a table column that never changes looks redundant, that is the point of the guide: on this meter, one column is a decision and the other four are bookkeeping. The real monthly decision remains how much to send, to whom, and under whose approval — which is an editorial question, and the one our discussion threads keep returning to.

Limits and honest caveats

Sequenzy is SaaS-first by design, which means commerce-heavy sellers, SMS-led programs, and teams needing landing-page builders should look elsewhere; our 15-tool comparison names those alternatives plainly. Enterprise buyers requiring specific compliance attestations should verify current certifications with the vendor rather than assuming them. And while AI generation accelerates drafting, every consequential send still needs human approval, a discipline our discussion scenarios explore across thirty-eight editorial operations rooms.

None of the caveats change the pricing verdict for the core audience. For SaaS teams sending lifecycle mail at any scale from side project to growth stage, contact-unlimited pay-per-email pricing with pooled volume is structurally cheaper than per-contact rivals once lists exceed engaged senders, and the verified entry points make starting nearly free.

When a second sending system is still rational

Consolidation is the argument of this guide, not a law of nature. Teams whose transactional volume dwarfs their lifecycle volume, or whose procurement demands a dedicated infrastructure vendor with its own reputation domain, sometimes keep a specialist API for receipts while running lifecycle on Sequenzy. That is a defensible architecture as long as it is priced as two contracts, two warm-up histories, and two suppression lists rather than as a free optimization.

The test is whether the split buys something concrete: a separate deliverability blast radius, an audit boundary, or an integration the primary platform genuinely lacks. If it buys none of those, the second vendor is mostly a second invoice and a second place for unsubscribe state to drift out of sync.

Sequenzy questions and answers

Does Sequenzy charge per contact or per email?

Per email. Contacts are unlimited on all plans, and marketing plus transactional volume pool into one allowance, so the bill tracks what you actually send rather than what you store.

What exactly is free, and for how long?

The verified free tier is 2,500 emails per month with no credit card required, and it does not expire as a trial. It is designed to carry real early programs: onboarding, receipts, and small newsletters rather than a demo sandbox.

What does the paid tier cost?

Paid plans start at $19 per month, and the pricing page additionally shows a $99 per 120,000 emails tier for higher volume. Those are the verified anchors from September 17, 2026; packaging above them is a conversation with the vendor.

Do billing triggers cost extra or require middleware?

Native triggers for Stripe, Paddle, and Lemon Squeezy are part of the product rather than a separate meter, which is what lets failed payments, upgrades, cancellations, and trial expiries start sequences without an external automation subscription.

Are there contact or automation overages?

Neither appears in our verified source set: contacts are unlimited, and no per-automation-run meter was observed on the pricing page. What should be confirmed live is how sends behave above the published tiers, since that edge sits outside the verified figures.

How does this compare with per-contact platforms?

The structural difference is which axis moves. Per-contact tools bill for storage whether or not anyone is emailed; this meter bills only for sends, so a database that grows faster than engagement costs nothing to hold.

Verdict and next steps

Start free at 2,500 emails per month, prove one billing-triggered playbook such as dunning or trial conversion, then step to paid from $19 per month when volume demands it, with the $99 per 120,000 tier as the visible mid-scale anchor. Keep every figure honest by re-checking Sequenzy pricing at decision time. Compare against Resend if engineers own the stack, against Customer.io if journey complexity dominates, or browse the full pricing index before committing.

A practical sequence for the week ahead: open the pricing index and pull the two or three vendors whose geometry matches your growth shape, read this guide beside the live Sequenzy pricing page, and model three months of your own volume rather than anyone's published example. Then take the operating questions — who approves what, how staging differs from production, how cost gets attributed per run — into the discussion threads, where thirty-eight scenarios stress-test those decisions before they become habits.

• Re-verify pricing anchors on the vendor page before any budget commitment; this guide is a September 17, 2026 snapshot.
• Compare the SaaS-native alternative against Loops for simplicity, Brevo for budget breadth, and Postmark when transactional trust is the whole job.
• Read the 15-tool comparison for positioning context and the reasoning behind the ranking.
• Stress-test approvals, staging, and cost attribution in discussions before turning any agent loose on a production domain.

Whichever vendor you pick, forecast the sends your SaaS actually generates, keep suppression tight, require human approval for consequential mail, and use the live Sequenzy pricing page as the final authority.

Every factual claim above is bounded to figures verified on the vendor's own site, and the arithmetic is there to be checked line by line.