In Mercor’s comparison, Claude Opus 5 scored 100% across 20 attempts at four simplified month-end-close scenarios and completed every attempt in under 10 minutes. The human comparison involved 12 licensed CPAs, but it tested structured, document-heavy exercises—not whether an AI can independently close a real company’s books. Mercor’s broader APEX-Accounting benchmark makes that limitation clear: models often earned partial credit, yet rarely completed tasks correctly and consistently across repeated runs.
What the CPA comparison actually tested
Mercor compared Claude Opus 5 with 12 licensed CPAs, whose average accounting experience was about five and a half years. Participants worked through four month-end-close scenarios adapted from APEX-Accounting. They searched company files, found figures, performed calculations, and returned results in tables. The AI scored 100% on all 20 attempts, with each attempt taking less than 10 minutes. The CPAs worked without AI assistance.
As an Amazon Associate I earn from qualifying purchases.
Mercor says it had planned to measure how AI might augment accountants. Because the AI achieved perfect scores on these simplified tasks, the experiment instead compared unassisted human and AI performance. The result is evidence that the model handled this narrow set of well-defined exercises quickly and accurately; it is not evidence that it outperformed accountants across the profession. Mercor’s October 1, 2026 write-up describes the limits of the comparison.
Why a perfect score does not mean an autonomous close
A month-end close is not just locating numbers and filling out a table. Accountants also work with people, ask questions when records or instructions are unclear, apply business context, and resolve exceptions. Those parts of the work were not represented in Mercor’s four scenarios. The comparison also did not test a live company’s books, deployment in a finance team, or whether an AI could safely approve and sign off on a complete close.
#1 Best Overall
- Profitability calculations; cash flow function Calculates NPV and IRR for uneven cash flows
- Time-value-of-money and Amortization keys solve problems including: pension calculations, loans, mortgages, etc.
- Ideal calculator for students, managers and statisticians
- Built-in functionality : List-based one- and two-variable statistics with four regression options: linear, logarithmic, exponential and power
- The BA II Plus calculator is approved for use on the following professional exams: Chartered Financial Analyst exam. GARP Financial Risk Manager (FRM) exam. Certified Management Accountants exam
Mercor researcher Aden Barton wrote that the tasks tested areas where AI is strong, such as detail-oriented instruction-following, rather than the whole accounting job, which includes less structured work such as client communication. That qualification matters: a model may produce a correct answer when the task and supporting files are clearly specified while still needing a person to decide whether the task is complete, whether the inputs make sense, and what to do when they do not.
The broader APEX benchmark tells a different, complementary story
The four-scenario human comparison should not be conflated with the separate APEX-Accounting benchmark. APEX covers 160 tasks across 10 fictional company worlds, created by more than 40 accounting professionals with a median of 11 years’ experience. In Mercor’s July 2026 report, the leaderboard leader, Claude Fable 5, scored 56.4% Mean Criteria@3. That is a different model snapshot, task set, metric, and evaluation from the Opus 5 comparison.
Rank #2
- Large 12 Digit Blue Fluorescent Display
- 2.5 LPS, 2 Color Ink Printer
- One-Touch Change Calculation Function
- Cost/Sell/Margin keys, Grand Total and Average
- Uses replaceable Victor ink roller IR40T
More importantly, partial credit is not the same as finishing an accounting task correctly. APEX reports that at least one model earned some credit on more than 95% of tasks, but 58% of tasks were never fully solved by any model on any run. The most consistent model fully solved only 2.6% of tasks in all eight runs. Mercor summarized the result this way: “The headline result is not which model leads, but how rarely any model succeeds consistently.”
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Those figures do not contradict the perfect result in the smaller comparison. The exercises and scoring conditions differ. Together they suggest that models can excel on bounded, structured work while remaining unreliable at completing broader tasks end to end. See Mercor’s APEX results and methodology and the APEX-Accounting technical report.
Rank #3
- Two-way Power Desk Calculator: Use solar power or battery power,In the case of sunlight or light, it can also be used without battery (Provide 2 AA batteries, only 1 needed).
- Optimized for Desk Use: The angled display offers a better viewing angle, especially when placed on a flat surface.
- Ergonomic Screen Tilt: Reduces neck strain with a user-friendly viewing angle, naturally aligning with your line of sight for a more comfortable experience.
- 10-Key Calculator with Large Buttons: Easy-to-use design follows computer keyboard layout.
- Desktop Basic Office Calculator:Perfect for daily use in offices, businesses, schools, retail stores, shopping centers, and home offices.
What the benchmark leaves outside its scope
APEX-Accounting uses fictional company environments and does not evaluate tax, audit, consolidation, multi-entity or multi-currency accounting, or external reporting. Its report also says the benchmark does not test how an agent responds when it needs clarification. Neither APEX nor the CPA comparison establishes productivity gains in deployed businesses or shows that an AI can manage a real close without human review.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to interpret the cost and speed figures
Mercor calculated a cost of $0.21 per rubric criterion for Claude Opus 5, compared with $10.35 per criterion for unassisted accountants, using the US median accountant wage for the human-cost calculation. This is a task-level comparison tied to the study’s rubric, not the total cost of owning or deploying an accounting AI system. It does not include the full costs or operational requirements of software, integrations, human review, exception handling, or a real finance workflow. The sub-10-minute completion time likewise applies to the study attempts, not to real-world close schedules. The Decoder’s October 2, 2026 coverage reports the cost comparison.
Rank #4
- Check your calculations thanks to the calculator's inbuilt serial impact roller printer that enables you to monitor your inputs and retains ongoing records. This two-color printer with a four-key memory prints red and black ink at up to 2.3 lines per second onto the included roll of paper.
- Printing calculator offers 12-digit LCD display for convenient viewing. 4-key memory keeps often-used figures accessible for faster calculations. Clock and calendar functions help maintain schedules.
- Easy-to-use solution for all of your basic math needs. Streamline financial calculations with currency conversion, tax calculation, and item counter functions. 1-year manufacturer limited warranty.
- Dimensions: 2.2"H x 6.4"W x 9.1"D. Package content: AC adapter, paper roll, user manual.
- Powered by any standard AC outlet, eliminating the need for expensive batteries. Decimal switch, rounding switch, percent, sign change, backspace, double zero, and grand total functions help you solve a variety of mathematical problems.
What businesses can reasonably take from the results
The evidence supports evaluating AI for bounded, document-heavy accounting tasks, with a qualified person reviewing outputs and handling ambiguity. It does not support removing accountants from a close or letting a model independently certify that books are complete. When comparing accounting AI claims, look beyond a headline score:
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchQuick Recap
- Task scope: Were the tasks simplified exercises, realistic workflows, or live-company work?
- Completion standard: Does the score reward partial criteria, or require a fully correct end-to-end result?
- Repeatability: Does the model succeed across repeated runs, or only on its best attempt?
- Evaluation conditions: Which model version, task set, tools, and budget were used?
- Human work: Were communication, clarification, business context, and review included?
- Cost basis: Is the number per benchmark criterion, or does it include the costs of operating and supervising a deployed system?
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
