Canopy AI: Smart Tax Prep, Document Renaming and Beta Features

Canopy, the practice-management platform for tax and accounting firms, offers Smart Tax Prep (draft returns from partner AI tools, reviewed by your team), AI Document Renaming and Matching, an AI Meeting Notetaker and Smart eSignature Fields. On its pricing page the notetaker, signature fields and Canopy Coworker are labelled Beta; Smart Tax Prep is labelled New.

A note on names before going further. "Canopy" is used by several unrelated companies and products. Everything here is about the practice-management software sold at getcanopy.com to accounting and tax firms. Canopy also sells other things in the same suite, such as Close Automation and Tax Resolution, which I keep separate below, because they answer different questions from "what AI helps with tax work".

Follow me on Instagram@sagnikteaches

The features at a glance, with the label Canopy gives each

I read Canopy's pricing page and its product pages in November 2026. The pricing page carried beta labels that its product pages did not repeat, so treat the labels as a prompt to ask, not as a settled fact.

Connect on LinkedInSagnik Bhattacharya

Plan prices as listed, per user per month billed annually: Standard $74, Plus $109, Premium $149 (marked coming soon), and a custom Enterprise tier. The page lists $93, $136 and $186 monthly for the first three, and says annual billing saves 20 per cent against monthly. The notetaker allowance is listed as 8, 12 and 16 hours per user per month on Standard, Plus and Premium. Tax Workflow Automation (Smart Intake and Smart Delivery) comes with 5 client credits on each plan, renewed each term, and extra credits start at $34 per client with volume discounts. Canopy's FAQ says a credit gives unlimited use for one client for the rest of the subscription term, and is used the first time the AI touches that client: an AI-generated questionnaire or document checklist, a questionnaire sent with "Pre-Fill with AI", or a tax delivery task. Questionnaires and checklists built and sent without AI use no credit.

Subscribe on YouTube@sagnikteaches

Those credits matter because of where the features sit. Canopy launched Tax Workflow Automation on 15 July 2026 and announced general availability on 3 November 2026, as three modules: Smart Intake (AI-built questionnaires and checklists, plus recognising, renaming and filing uploads), Smart Prep (the partner hand-off described below) and Smart Delivery, which reads the finished return and builds a flow where the client reviews it, signs and pays, with the option to hold the return until the invoice is paid. Canopy's releases quote time savings per client and per return. Treat those as the vendor's figures and measure your own.

Smart Tax Prep: what it is and what it is not

Smart Tax Prep is the feature most people mean when they ask about Canopy AI for tax, and the most often misunderstood. Canopy describes it as a partner-agnostic framework for AI-powered tax preparation. Documents sync from Canopy to an AI partner, the partner extracts, classifies and maps the data and generates a draft return, and the status, errors and completed workpapers show up in your Canopy work lists. Your team reviews and approves each return before it is finalised.

Three details matter for a practice owner.

• Partners, not an in-house engine. The Smart Prep page names Filed, Byron and Juno as partners (the pricing page mentions only Filed and Juno, so confirm who is current), and a firm picks the one that fits its return types. Each partner requires its own subscription or plan, priced by partner and by the number of returns processed. So the cost of Smart Tax Prep is the partner's bill, not just your Canopy plan.
• Works beside your tax software. The page says it works alongside tools such as Drake, CCH Axcess, ProConnect or UltraTax, rather than replacing them. The return still ends up in your preparation software.
• Scope is narrow on the page. The only return type Canopy names is the 1040, and it says the feature removes manual data entry for them. Do not assume business, trust or state returns without asking.

Canopy also states that data sent to partners is encrypted in transit and at rest, and that integration partners must meet SOC 2 standards. That is the vendor's statement. A firm with its own duty of care for client data should still ask each partner for its security documentation and where processing happens. The comparison of TaxDome, Canopy and ChatGPT for tax preparation puts this in the wider tool choice.

A sample walk-through, with illustrative numbers

Say a five-preparer practice handles 400 individual returns a season. A preparer currently spends about 35 minutes per return keying data from W-2s, 1099s and bank statements. The practice runs a pilot of 20 returns through Smart Tax Prep with one partner. Illustratively, the pilot shows these things.

• Of 20 returns, 14 drafts need only light correction, 4 need substantial correction because a document was misread, and 2 fail with an error status and go back to manual.
• For the 14, review takes about 15 minutes instead of 35. For the 4, review and correction take about 40 minutes, longer than manual work. The 2 failed returns cost five minutes of wasted upload time each.
• Total pilot time: 14 x 15 = 210, 4 x 40 = 160, and 2 x 35 (manual after failure) + 10 wasted = 80, so 450 minutes against 700 manual, about 36 per cent less.

Then add the partner's per-return fee, the time to upload and fix scans, and the senior review that the saved minutes do not replace. Only after that sum does the saving show up. If the saving is real but small, the feature may not justify the partner's bill. The pilot plan for AI tax software sets out how to measure cost per return properly.

AI Document Renaming and Matching

This is the unglamorous feature, and often the one that pays back first. Canopy's description: automatically rename files and classify them by type or issuer under global rules, and auto-match uploads to checklist items. Canopy Coworker also checks new uploads against prior years to flag missing items.

The firm's naming rules decide the result. Without rules, you get tidy-looking chaos. Say a practice decides every file is named in the pattern ClientCode_Year_Type_Issuer.

Before (what clients upload)        After (renamed under the rule)
IMG_4471.jpg                        SMI014_2025_1099-INT_Northbank.pdf
scan0003.pdf                        SMI014_2025_W-2_Brightwell-Ltd.pdf
Document (7).pdf                    SMI014_2025_Mortgage-Interest_Harbour-Loans.pdf

That kind of result is what the feature aims at. In a test you check three things. How many of 100 uploads are classified correctly. How many are matched to the right checklist item. And what happens with the ones it cannot read: does it leave the original name and flag it, or does it guess? A silent wrong guess is worse than an honest "unrecognised".

Say the practice tries it on 100 uploads from last season and 88 are renamed correctly, 7 are flagged as unclear, and 5 are wrong, for example a brokerage statement labelled as a bank statement. The 5 wrong ones are the worry. The practice makes a rule: a person looks at every file the AI has renamed before a preparer opens the folder, until two weeks of spot checks show no errors. The saved time still adds up: a team spending 10 hours a week sorting files may save most of it.

AI Meeting Notetaker

Canopy says the notetaker converts meeting transcripts into notes saved to the client's file. The pricing page labels it Beta and lists monthly hour allowances by plan. For a tax practice it matters in two seasons: planning meetings in the off-season and review calls at filing time.

Before you switch it on, settle four practical points.

1. Consent. Tell clients that a meeting is recorded and transcribed, and get any consent your rules or clients' expectations require. A line in your engagement letter and a spoken reminder at the start of a call is a common approach.
2. Where the notes go. Notes saved to the client file are a record. Decide who can see them and how long you keep them.
3. Accuracy on numbers. Transcripts get figures wrong. Rule: any figure or date in an AI note is checked against source before it is relied on.
4. Allowance. The pricing page lists 8, 12 or 16 hours per user per month by tier, so the question is who runs the meetings. A practice holding 20 client meetings of 45 minutes a month uses 15 hours in total, comfortably inside five users' allowances. But if one partner runs 14 of those meetings, that is 10.5 hours on one login, past Standard's 8. Ask Canopy what happens at the limit and whether hours can be shared.

A short illustrative note shows both the value and the risk. Suppose the tool produces: "Client plans to sell the rental property in March; discussed capital gains; client will send the purchase paperwork." A good summary. If it adds "estimated gain $40,000" and nobody said it, that is the invented-number failure to watch for. The comparison of AI note takers for accountants covers cost and retention across tools.

Smart eSignature Fields

Canopy lists this as AI that auto-places signature fields in documents. eSignature itself is part of document management on every plan, which Canopy says includes unlimited eSignatures; identity checks before signing (knowledge-based authentication, KBA) cost $1.25 per credit, with volume discounts. The "smart" part is the placement: you upload an engagement letter or an e-file authorisation form, and the AI puts signature, date and initial fields where they belong.

The test is simple, and worth doing on your own forms, not Canopy's samples. Upload ten different documents: engagement letter, authorisation form, an organiser, a multi-page agreement with two signers and a PDF scan that is slightly crooked. Count how many fields are placed correctly, how many are missing, and how many land on the wrong signer. Because the feature is labelled beta, plan for corrections, and never send a document for signature without opening it first to check the fields.

Other Canopy products, kept apart from the AI question

Two paid add-ons turn up on the same pricing page and do not belong in a list of Canopy's tax AI. Close Automation is priced at $10 per connected client per month, with the first five connections included. Tax Resolution is $50 per user per month, with volume discounts. Both are separate products with their own purpose. Leave them out of a Smart Tax Prep comparison, and ask about them separately only if the practice does bookkeeping close work or tax resolution work.

Close Automation does use AI (Canopy says it scans the ledger for issues) and launched in open beta in April 2026, so a firm with monthly bookkeeping clients may want to test it. Test it as a bookkeeping product on its own timetable, not as part of the tax-season decision.

What the plans cost a small practice, worked through

Take a five-user practice on Standard billed annually. Five users at $74 a month is $370 a month, or $4,440 a year. On monthly billing the listed $93 a user is $465 a month, $5,580 a year, so the annual route saves $1,140. Move everyone to Plus at $109 and the annual figure is $6,540. Premium is marked coming soon on the page, so do not plan around it.

Then add what is not in the plan price. Smart Tax Prep partners bill by return, so if a partner charged an illustrative $8 a return and you pushed 200 returns through, that is $1,600 a season on top. Tax Workflow Automation credits are per client: a practice that wants AI intake and delivery for 200 clients needs 195 credits beyond the five included, which at the $34 starting price is $6,630 a term before any volume discount. If one person runs most of the client meetings and needs more notetaker hours, the answer may be a higher tier for that user only. None of these numbers is a quote, and each depends on a conversation with Canopy and the partner.

The test that matters is the one at the end of the pilot: hours saved, multiplied by what an hour of preparer or admin time is worth to the firm, set against the extra cost. Say the practice saves 120 hours over a season and values an hour at $45 for internal cost. That is $5,400. If the added spend, partner fees plus the plan difference, comes to $3,000, the feature earns its place. If it comes to $6,000, it does not, however good the demo looked.

Which feature suits which kind of practice

A seasonal individual-returns practice

The firm lives or dies by file volume in February and March. Document renaming and matching and Smart Tax Prep are the candidates, because both attack the keying and sorting that eat the season. The notetaker matters less, since meetings are short and numerous. Test document renaming first: it is cheap to measure and needs no partner contract.

An advisory-heavy practice with fewer, larger clients

Here the work is conversations. The notetaker is the candidate, with consent handled properly, and Smart eSignature Fields help with engagement letters and authorisations. Smart Tax Prep may matter little if most returns are complex and not 1040-only.

A bookkeeping-plus-tax firm

The monthly cycle matters more than the season. Close Automation is a separate add-on at $10 per connected client per month, with the first five included, so it is a separate sum. For the AI question, document matching against checklists and the notetaker may bring more than Smart Tax Prep.

Misunderstandings that cost a pilot its credibility

• "Beta means it is free to try." A beta label says nothing about price. Check the plan inclusion in writing.
• "The AI drafts, so review can be light." Review of an AI draft needs as much care as a new junior's work, because the errors are plausible. Set the review standard before the pilot.
• "Canopy AI" means one thing. It is five or six features with different labels, partners and costs. Judge each separately.
• "Our tax software will import it automatically." The draft lands in your Canopy work list. Confirm how the data reaches Drake, CCH Axcess, ProConnect or UltraTax, and whether anything is re-keyed.
• "The labels are final." I saw beta labels on the pricing page and none on the product pages. Ask Canopy which is current.

A two-week plan to test the beta features without risking the season

1. Days 1 to 2: write the questions. For each feature, write the one number you want to move: minutes per return, files correctly named, notes accepted without edit, forms sent without field fixes.
2. Days 3 to 7: run last year's work. Use closed files, not live ones. Document renaming and Smart Tax Prep are easy to test on past clients, and the answer is already known.
3. Days 8 to 10: run a small live batch. Five to ten clients who agree, with a person reviewing every output.
4. Days 11 to 12: count. Time saved, errors caught, errors missed, and anything that failed silently.
5. Days 13 to 14: decide. Keep, expand, or switch off. Write the decision and the reasons in one page, and set a date to look again.

Run the pilot outside the busiest weeks. A beta feature that stumbles in the middle of filing season costs more than the same stumble in the autumn. If your firm is deciding between platforms rather than testing features, the Canopy versus TaxDome tutorial frames the choice by season.

Questions to put to Canopy in writing

• Which of these features are beta today, and what does beta mean for support, change notice and data handling?
• Which plans include Document Renaming, Smart eSignature Fields and Smart Tax Prep?
• What happens when a user reaches the notetaker's monthly hours, and can hours be shared across users?
• Does one client credit cover both Smart Intake and Smart Delivery for that client all term, and what volume discount applies at our client count?
• For Smart Tax Prep: which return types are supported beyond the 1040, and what does each partner charge per return?
• Where is client data processed and stored, and can we have the partners' security reports?

Keep the answers with your vendor file. Labels and plan inclusions change often, and a written answer from the date you bought protects you. If you want help choosing which of these features deserves a pilot in your own practice, my AI implementation consultation includes mapping the tasks and setting the measures. For a view of how AI fits a tax practice across the whole season, see the tutorial on small tax practices using AI through the busy season.

Canopy AI questions from tax practices

Is Canopy Smart Tax Prep a tax engine?

No. Canopy describes it as a framework that sends documents to separate AI partners, which draft the return for your team to review and approve. It works alongside your existing tax software rather than replacing it.

Do I pay extra for Smart Tax Prep?

Canopy says each AI partner requires its own subscription or plan, priced by partner and number of returns. The Canopy plan itself is separate. Ask for a written quote from the partner you choose.

Which Canopy AI features are in beta?

At the time of checking, Canopy's pricing page labelled Canopy Coworker, the AI Meeting Notetaker and Smart eSignature Fields as Beta, and Smart Tax Prep as New. Its product pages did not repeat those labels, so confirm each one with Canopy.

Can the AI file a return for me?

No. Canopy's description has your team reviewing and approving each return before it is finalised. The preparer remains responsible for the return, and the AI output is a draft.

Is the notetaker safe to use with client meetings?

Tell clients you record, get any consent your rules require, and check where recordings and transcripts are stored and for how long. Ask Canopy for its data terms in writing before you use it with sensitive meetings.

Further reads

• TaxDome Pricing: Plans, Seasonal Seats and What to Ask on the Demo — What to ask on a TaxDome demo and how seats are priced.
• TaxDome vs Karbon for a Small Firm: AI, Cost and Fit — AI, cost and fit for a small firm choosing between two tools.
• AI Tax Prep Software for CPAs: What to Test Before You Switch — What to test before switching AI tax prep software.
• AI Receptionist for Accounting Firms: Tax-Season Calls and Cost — Tax-season call handling, with cost.
• AI Tax Research Tools Compared: Citations, Cost and Checks — Tax research AI tools compared on citations and cost.
• Canopy Pricing: Power-Ups, Client Credits and Add-On Costs — Canopy's tiers, tax workflow credits, Close Automation and payment fees priced out for four firm shapes, with the questions to ask before you sign.
• AI Tax Return Prep for Small Practices: Extract, Review, Sign Off — A three-stage workflow for AI-assisted return prep: extract with page references, tie out with a spreadsheet, then review and sign off in writing.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Canopy pricing page, Smart Prep page, Tax Workflow Automation page, document management page and press releases of 15 July 2026 and 3 November 2026 (getcanopy.com, checked November 2026). Labels and plan inclusions change, so confirm with Canopy.