HVAC Takeoff Software: 6 Steps to Choose in 2026

Pick a quantity tool or a duct-labor tool first

HVAC takeoff software is the program an estimator uses to measure duct, pipe, equipment, and air devices on a drawing and turn those measurements into quantities a bid can price. The decision is not which highlighter feels fastest on a clean sample sheet. It is whether this bid needs a duct-labor engine, a quantity tool that leaves an audit trail, or both. A linear foot of supply main is not a price. Pressure class, fittings, and the labor table behind them are the price. Length, area, and count still matter, because someone has to show where the quantity came from, but they do not replace a sheet-metal estimate.

TL;DR: If the number you send has to carry duct labor, start with a duct estimating program and use a measurer to check the drawing. If you already price labor in a workbook and you need counts, lengths, overlays, and a file a second person can open, shortlist the published-price tools below and reject any plan that cannot count or that caps the sheet set under your real jobs.

Shops are making this call while hiring stays tight. Craft openings: 87% of firms according to AGC (2026). A seat that an estimator can run without a week of custom setup is part of how a short staff keeps bidding. It is not a substitute for a reviewer who knows the spec.

Key Takeaways

• The first cut is duct labor versus drawing quantities. A length tool does not price pressure class, fittings, or install hours.

• Read the billing unit before you compare seats. One published price is per year. Another is per month, billed annually. A third is per market and is not takeoff at all.

• On Bluebeam, length and area start on the cheap plan. Counts and perimeter start higher. The Excel quantity link starts higher still.

• ConstructConnect Takeoff names interior trades and caps the cheaper takeoff seat at 25 plans. The project-search prices on the same screen are a different product.

• An AI first pass still needs a person on scale, missed devices, and the addendum before anyone multiplies the quantity by labor.

• A handoff past the takeoff file is worth specifying only after an export or API exists and a named person accepts the change list.

Who this is for

This guide is for an estimator at a commercial mechanical or HVAC contractor who is close to choosing. You already bid from PDF or CAD. You care whether the tool can count diffusers and VAV boxes, measure duct runs, survive an addendum the afternoon of the bid, and hand a file to someone who did not draw the takeoff. You may also price sheet-metal labor, piping, or both, and you need to know which programs even claim that job.

It is also for the estimating manager who has to explain the seat to accounting. Published prices below are copied from vendor pricing pages opened for this guide. Where a vendor does not print a license price, the cell says Quote-based. Nothing here is a timed study of your drawings, and no customer result is implied.

Red flags: you only mark PDFs for a general contractor and you will never price labor; nobody on the team will review quantities before they hit the proposal; your sets routinely run past a 25-sheet cap and you are only considering plans with that cap.

The office still has to catch the general contractor while you are inside a set. That phone problem is a different buy from takeoff, covered in answering services for HVAC contractors. Takeoff also does not track cylinders or leak logs. That record lives in refrigerant tracking for HVAC contractors.

HVAC technicians: 77% hard to fill according to AGC (2026). Field hiring does not pick your software. It does explain why a missed device count or a wrong pressure class is expensive: the people who would catch it in the shop are already scarce.

How we evaluated these tools

The weights below are a buying method for a commercial mechanical estimator. They are not vendor scores, not a survey of users, and not a lab test. A tool can be a strong PDF measurer and still fail the duct-labor weight. That is a scope mismatch, not a quality slur.

Growing firms: nearly 3 in 4 according to MCAA (2026). Three in four contractors in that polling describe themselves as moderately to highly engaged with technology, and more than half say their process documentation is insufficient, according to MCAA (2026). Growth plus thin documentation is why the export and the human accept step carry real weight. A faster first pass that nobody can retrace becomes a bid you cannot defend.

Seven products stayed in the set because each has a live product page and is sold now: QuoteSoft, Bluebeam Revu, PlanSwift, STACK, On-Screen Takeoff, Togal.AI, and ConstructConnect Takeoff. QuoteSoft is the one built for duct and piping estimates. The others are quantity, markup, or general takeoff tools that an HVAC estimator might already have on the desk. Features in the matrix are limited to what those pages state. "Not stated" means the page reviewed did not say, not that the feature is impossible.

Published prices, side by side

Pricing checked October 14, 2026.

Compare the unit, not just the digit. A per-year desktop seat and a per-month cloud seat are not the same cash plan. Project-search prices that are charged per market are not takeoff seats, even when they sit on the same screen.

Core seat: $330 per year according to Bluebeam (2026). Basics is $260 per user billed annually, Complete is $440, and introductory Max is $590 per user billed annually, with that introductory Max price locked through the 2027 renewal for buyers who take it, on the Bluebeam pricing page. Basics measures length and area and does not include perimeter, counts, or the right to host a Studio session. Core adds perimeter, counts, angles, volume, overlay, and hosting. Complete adds Dynamic Fill, Quantity Link into Microsoft Excel, batch link and slip-sheet tools, and scripting. Max adds the AI drawing tools. All four include phone and email support, Bluebeam University, and Revu for Windows plus the web and mobile apps.

PlanSwift's Essential plan costs $2,000 per year for one seat, the Core plan costs $3,000 per year for one seat, the trial lasts 14 days with no credit card, live training is 2 hours, the page says the product is trusted by over 43,000 construction pros, additional seats are discounted with no dollar amount printed, and 10 or more licenses may qualify for a volume quote, according to PlanSwift (2026). Essential includes Auto Count, Auto Scale, and Auto Bookmark. Core adds Auto Takeoff for a first pass of areas, lengths, and counts. Both are annual plans covering updates and support for the year.

The Premium plan on STACK costs $299 per user per month, billed annually, the Pro plan costs $349 per user per month, billed annually, the free account is $0 with 1 active project a month, up to 300 plan pages, and 50 measurements per project after 7 days of Pro, the banner says save 15% on annual billing, and the FAQ says paying the year upfront saves 20% while monthly payments are also offered on a 12-month term, according to STACK (2026). Confirm the invoice. Those two discount lines do not match. Premium adds unlimited projects, assemblies, overlay, symbol search, and AI takeoffs. Pro adds the worksheet estimate, Excel, a regional cost library, and proposals. ERP connectors are an add-on on Pro.

Togal's Growth plan costs $299 per user per month, billed yearly, and the Business plan is custom for teams of 4 or more users, according to Togal (2026). Growth lists unlimited automated takeoffs, unlimited chat prompts, unlimited symbol search, and collaboration with people outside the account who do not need their own seat.

The Starter plan for ConstructConnect Takeoff is listed at $100 per seat per month and at $1,200 per year, with 1 seat and 25 plans per project, and the Professional plan is listed at $183 per seat per month and at $2,200 per year with no 25-plan cap, while project-search Starter is $129 per month per market, project-search Professional starts at $199 per month per market per seat when billed annually, and project-search Premium starts at $10,000 per year, according to ConstructConnect (2026). Takeoff seats are annual contracts. The pricing FAQ says mid-term cancellations are not refunded. A 14-day trial is offered and no license price is printed, according to On Center (2026). QuoteSoft's duct and pipe pages do not print a license price either, so both stay Quote-based.

Use this as an illustration, not a timed study and not a published labor rate: an estimator counts 120 diffusers at 15 seconds each, which is 30 minutes, then traces 2,400 feet of supply duct at 100 feet per minute, which is 24 minutes, then spends 18 minutes splitting one revised sheet, for 72 minutes before labor is priced. Spread the Core plan price of $330 per user per year on the Bluebeam pricing page across 110 bids and the seat is $3 a bid. If a missed pressure class on those 2,400 feet were priced at an illustrative $1.50 a foot, the miss would be $3,600, so a proposed handoff should pull the markup only with a valid access_token as described in Bluebeam's authentication guide and stop until a person accepts the lengths.

Six steps that keep a mechanical bid honest

1. Separate labor from quantity. Write down whether the bid must carry fabrication and install hours. If yes, a PDF measurer is the check, not the estimate. QuoteSoft is the product in this set that describes duct shapes and pressure-class labor. If the labor already lives in your workbook, you are buying measurement, overlay, and an export.

2. Reject plans that cannot count. Diffusers, grilles, and equipment are counts. A plan that stops at length and area will not finish a mechanical sheet. On Bluebeam that line sits between Basics and Core. On ConstructConnect Takeoff, Auto Takeoff is listed on Professional, not on Starter.

3. Read the unit on the price. Put the billing phrase next to the digit: per user per year, per user per month billed annually, per seat per year, or per market. Then check the cap. Starter on ConstructConnect Takeoff stops at 25 plans and one seat. STACK's free account stops at 50 measurements. Either cap can silently drop scope on a real package.

4. Test an addendum, not a clean sheet. Ask the vendor, or use the trial where one is published, to overlay a revised mechanical sheet. On-Screen Takeoff describes removed work in red and added work in blue. Bluebeam Core and STACK paid plans both describe overlay. If the only demo is a single tidy floor, you have not seen bid day.

5. Name the person who accepts an AI first pass. PlanSwift Core, STACK paid plans, Togal Growth, ConstructConnect Takeoff Professional, and Bluebeam Max all describe some automatic measure. None of those pages make the model the signer of the bid. Assign a reviewer before you turn the feature on. The accept step is a name on a checklist, not a hope that the count "looked fine."

6. Connect the export only after that acceptance. Trigger: the estimator saves a CSV or Excel quantity file after the overlay. Action: a proposed US Tech Automations workflow compares that file with the last accepted version and opens a review list of changed counts and lengths. Output: a change list, not a priced proposal. Prerequisite: the takeoff product must write a file you are allowed to read, or offer an API your account can call. A person still accepts the list before any labor rate is applied. This is a configuration you would specify. It is not a live deployment, and it is not a measured saving.

What each product is actually for

QuoteSoft. Best fit is a commercial sheet-metal or mechanical contractor whose bid is won on fabrication and install hours. The duct page describes rectangular, round, and oval systems, adjustable labor rates, pressure-class costs, and graphical, 3D, and automated takeoff from digital plans. A separate pipe product covers plumbing and piping, and the duct page names Autodesk Fabrication and CAMduct. The FAQ names QuickBooks among accounting connections. Limitations: no license price is published, and the pages reviewed do not describe a browser app, a sheet cap, or an overlay. Implementation is a quote plus setup of your specs and labor tables. That is slower than a card checkout, and it is the point if those tables are the company standard. Evidence is the QuoteSoft product page.

Bluebeam Revu. Best fit is a shop that already lives in PDF markups and needs scale, counts, overlay, and a path into Excel. Core is the first plan that can count devices and measure perimeter and volume. Complete is the first plan with Quantity Link. Max is the AI tier, at the introductory price above. Limitations: nothing on the pricing page is a duct catalog or a pressure-class labor engine, and Basics will stall a mechanical count. Web and mobile measurement is narrower than desktop Revu. The developer portal, including the access_token flow, is limited to active customers and partners in the US, AU, DE, UK, and SE. Implementation is per user, and plans can be mixed. Evidence is the Bluebeam pricing page.

PlanSwift. Best fit is an estimator who wants takeoff and estimating in one desktop-style annual seat, with assemblies, and who will pay for Auto Takeoff on Core rather than count symbols only. The trial covers PDF, JPG, TIFF, and CAD. Limitations: the pricing page does not describe duct shapes, pressure class, or overlay, and it does not print the extra-seat dollar amount. Teams of 10 or more are back in quote territory. Implementation includes the 2 hours of training on the pricing page, then your own assemblies. Evidence is the PlanSwift pricing page.

STACK. Best fit is a cloud takeoff that several people can open, with Pro if the estimate worksheet and Excel link have to live in the same subscription. Paid plans add overlay, symbol search, and AI takeoffs, plus unlimited viewer seats on the pricing page. Limitations: free is a sandbox, not a bid seat, once the 7 days of Pro end. The 15% banner and the 20% FAQ line need a written quote before you budget. Duct labor and pressure class are not described. ERP is an add-on, not part of the seat. Implementation is a 12-month term, annual or monthly payments. Evidence is the STACK pricing page.

On-Screen Takeoff. Best fit is a Windows estimator who thinks in conditions, typical areas, and repeating pages, and who wants a patented multi-condition pass plus a red-and-blue overlay. Auto Name and pixel auto-count are on the product page, and Takeoff Boost is offered as an AI preview. Limitations: no license price is published, the system requirements are Windows, and the page does not describe a duct labor database. Implementation is a 14-day trial, then a quote, with a local install and an activation code tied to the version. Evidence is the On-Screen Takeoff product page.

Togal.AI. Best fit is a browser-first team that wants an AI area and symbol pass, internal collaboration, and the right to invite outsiders who do not buy a seat. Growth is the published plan. Business, at 4 or more users, adds onboarding, a named support contact, a classification library, and SSO, on a quote. Limitations: the pricing page does not describe duct shapes, pressure-class labor, overlay, or Excel. Cancel by contacting them before renewal. Implementation is a credit card on Growth, or a sales conversation on Business. Evidence is the Togal pricing page.

ConstructConnect Takeoff. Best fit is a browser takeoff for linear, area, and count, with Auto Name and Auto Scale, when the work matches the trades the product page names: drywall, acoustical ceiling tile, painting, and flooring. Professional adds Auto Takeoff, unlimited plans, more than one paid seat, edit rights for collaborators, and up to 3 quantity results per condition, with CSV export. Starter is one seat, 25 plans, and no Auto Takeoff in the feature list. Limitations: it is the wrong primary tool for duct fabrication, and the project-search prices on the same screen will blow the budget if someone buys them thinking they are takeoff. Annual term, no mid-term refund stated. Evidence is the ConstructConnect Takeoff page and the pricing screen cited above.

When NOT to use US Tech Automations

Stay on the takeoff product you already trust when one estimator produces the quantity and the only chore left is pasting a total into the proposal. Skip an extra handoff layer when the vendor will not give you an export or an API, because there is nothing to move. Skip it on a one-sheet service bid that already fits a spreadsheet. A simpler file beats a second system when nobody is assigned to review the output.

Mistakes that show up on bid day

Buying the cheapest Bluebeam plan because the word "measure" is on it, then discovering counts are blocked, wastes the trial week. The fix is to point at one VAV schedule during the demo and confirm the plan you will actually buy can count it.

Treating the $129 or $199 project-search figures as the takeoff price mixes two products that share a website. Takeoff is the per-seat block with the 25-plan line. Project search is per market.

Trusting an auto count on a scanned sheet with no scale check pushes a bad length into every fitting. The reviewer should calibrate one known dimension before accepting the pass.

Pricing duct from a single linear-foot total, with no pressure class, underbids the heavy spec and overbids the light one. If the tool does not carry that factor, the labor table has to, and the export has to land in the workbook that holds it.

Letting a free or starter cap truncate the set is worse than a slow takeoff. Fifty measurements, or 25 plans, can drop a shaft or a revised floor without an error message you will notice at 4 p.m.

Sending the AI file onward before a person accepts it makes the handoff look tidy and the bid wrong. The accept step is the control. The software around it only moves what that person signed.

Glossary

• Takeoff: the measured quantity of a scope, taken from the drawings, before it is priced.

• Condition: a named thing you measure, such as a duct size or a device type, so later quantities stay grouped.

• Assembly: a recipe that turns one measured quantity into the parts and labor that go with it.

• Pressure class: the spec rating that changes gauge, reinforcement, and labor on duct. A length tool does not apply it unless you do.

• Addendum overlay: a comparison of two sheet versions so added and removed work is visible before you reprice.

• Quantity link: a live or exported tie from measurement markups into a spreadsheet. On Bluebeam it is a Complete-plan feature.

• Seat: one paid user. Viewer seats, where a vendor includes them, are not the same as estimating seats.

• Auto takeoff: a vendor's first-pass measure of area, length, or count. It is a draft until a person accepts the scale and the scope.

A no-code stack versus a configured handoff

Most estimating teams already own the alternative to a packaged handoff. They drop a CSV into a shared folder, or they wire Zapier, Make, or n8n to copy a column when a file appears. Those tools can keep a run history, retry a failed step, branch on a blank column, and store an audit trail when someone configures them. That is real capability. The buyer then owns idempotency so the same quantity is not posted twice, plus escalation when the file is late, access control on the workbook, and maintenance when a column header changes after an update.

Trigger: the reviewer marks the change list accepted. Action: US Tech Automations maps the accepted counts into the estimate sheet and holds the draft if the source call has no valid access_token, which Bluebeam's authentication guide requires on API requests. Output: a draft tab waiting on a second person before labor rates are applied. Prerequisites are an export or an API your account can call, a place to store the credential, and a named reviewer. This is a proposed configuration, not a deployment and not a measured result.

A proposed US Tech Automations design can set that accept gate, and the stop when the credential is missing, as the default path rather than a flow each estimator builds. The same prerequisites remain. The same person still has to know the drawing. Sitting that layer above tools you already run is a separate decision from the seat itself, which is the point of what changes when software agents sit above existing tools.

Questions estimators ask

Is the Basics plan on Bluebeam enough for an HVAC count?

No. Basics measures length and area and does not include counts or perimeter, which is where devices and many duct runs sit. Core is the first published plan that adds those measurements, and Complete is the first that links quantities into Excel.

Do On-Screen Takeoff and QuoteSoft show a license price?

No. Both are Quote-based on the pages reviewed here. On-Screen Takeoff does publish a 14-day trial. QuoteSoft describes the duct and pipe products and the labor and shape coverage, then sends you to a sales conversation for the number.

Is ConstructConnect Takeoff a duct fabrication program?

No. The product page names interior trades: drywall, acoustical ceiling tile, painting, and flooring. It measures linear, area, and count in the browser, which can help with devices and areas, and it is the wrong primary system when the bid is rectangular, round, and oval duct with pressure-class labor.

Can the free STACK account carry a real mechanical bid?

No. After the short Pro preview, the free account is capped at one active project a month, a few hundred plan pages, and a few dozen measurements. Those caps are in the pricing table. A commercial package will hit them.

Should an AI first pass go straight onto the proposal?

No. Use it as a draft. A person checks scale, missed symbols, and the addendum, then accepts the quantity. Only then should labor rates be applied. The vendors describe speed. They do not describe the model as the signer of your bid.

When does Zapier, Make, or n8n beat a configured handoff?

When one estimator already produces a trusted file and the only job is to move it, retry it, and log it. Those tools can do that if you design the checks and you keep them. They are the wrong save when you also need a named accept step and a stop on a dead credential, and nobody is going to own that build.

If the shortlist is a duct program plus a quantity tool, price the labor tool first and keep the measurer as the audit trail. For the handoff after a person accepts the quantities, see how US Tech Automations configures this.