MCP App Store
Obolus icon

Obolus

Open in ChatGPT

Overview

Obolus helps users calculate net income from gross salary and compare salary, tax, and social contribution outcomes across supported countries. The app supports detailed payroll calculations for one or two persons, cross-country net income comparisons, and a ChatGPT result panel for relocation, job offer, and salary planning workflows.

Tools

berechne

ChatGPT
Use this only for a detailed country-specific payroll or net salary calculation for one country/person workflow. For multi-country salary, net income, tax burden, or social-contribution comparisons, use taxcompare first. berechne is country-polymorphic: the same field names can mean different things depending on Land and the selected tax system. Use the declared Obolus field names exactly: top-level Land, Stjahr, Currency, LZZ, Modus, Personen. LZZ maps to 1=year, 2=month, 3=week, 4=day, and salary amounts such as Personen[].Gehalt_ct must be the gross amount for that selected period in minor currency units. Country-specific usage: DE: German payroll. Steuerklasse 1-6; Bundesland is the full Obolus state name such as Berlin or Nordrhein-Westfalen. For normal employees use the MCP preset de_payroll_preset=DE_STANDARD_EMPLOYEE_STATUTORY plus de_health_extra_contribution_percent as a literal percent such as 2.5. ChatGPT should not fill raw Obolus German payroll flags; the adapter translates semantic presets and aliases into backend compatibility fields. AT: Austrian payroll. Steuerklasse 1=standard employee, 2=sole earner with child, 3=single parent with child; Bundesland is the Austrian federal state; annual/monthly periods are the normal UI paths; pendler fields are Austria-only commuter inputs; BK/STS/STX/NETX can represent 13th/14th salary handling. CH: Swiss payroll. Steuerklasse 1=Tarif A, 2=Tarif B, 3=Tarif C, 4=Tarif H; Bundesland is the canton, currently Zurich in the UI; payroll deductions can include AHV/IV/EO, ALV, occupational pension, accident/sickness items; private KVG/LAMal health bills are not modeled as payroll health insurance. US: United States payroll. Steuerklasse 1=Single, 3=Joint, 4=Head of Household; Bundesland is the state; PreTax_Deduction_ct is for 401k/IRA/HSA-style deductions; Sonstige_Bezuege is supplemental income; outputs can split federal/state tax and OASDI/Medicare-style payroll contributions. UK: United Kingdom payroll. Steuerklasse 1=1257L standard tax code; Bundesland is region; PreTax_Deduction_ct is salary sacrifice or pension; Hat_Studienkredit toggles student loan repayment; outputs include PAYE income tax, National Insurance, and student loan where applicable. IE: Ireland payroll. Steuerklasse 1=Single, 2=Joint one income, 3=Joint two incomes; PreTax_Deduction_ct is pension contribution; Hat_Studienkredit is only relevant when a student-loan deduction is modeled; outputs include PAYE, USC, PRSI, and tax-credit information. CA: Canada payroll. Steuerklasse represents the federal/provincial claim setup; Bundesland is province such as Ontario, Quebec, or British Columbia; PreTax_Deduction_ct is RRSP/pension-style deduction; outputs can include federal/provincial tax, CPP/QPP, EI, and RRSP/credit information. AU: Australia payroll. Steuerklasse 1=Resident, 2=Non-Resident, 3=Working Holiday; PreTax_Deduction_ct is super salary sacrifice; Hat_Studienkredit toggles HELP/HECS repayment; Sonstige_Bezuege is bonus pay; outputs include PAYG tax, Medicare Levy, voluntary super, and employer super info. For standard German employees, ChatGPT/MCP should use semantic preset fields, not legacy numeric insurance fields: de_payroll_preset=DE_STANDARD_EMPLOYEE_STATUTORY, de_statutory_pension_and_unemployment=true, de_statutory_health_and_care=true, de_health_extra_contribution_percent=2.5 for a 2.5% Zusatzbeitrag, Kirche=0, KinderPVA=0, Kinderfreibetrag=0. Do not encode de_health_extra_contribution_percent as 25, 250, basis points, tenths, or cents. Raw Obolus fields such as gesetzliche_RV, gesetzlicheKvPvStatus, KV_Art, and KVZ are compatibility fields that can be valid in existing product/API payloads, especially for pensioners, civil servants, pension-exempt workers, and private health insurance. ChatGPT-facing calls should prefer semantic preset fields when possible. For "Germany, normal employee, tax class I", send de_payroll_preset=DE_STANDARD_EMPLOYEE_STATUTORY, de_statutory_pension_and_unemployment=true, de_statutory_health_and_care=true, and de_health_extra_contribution_percent=2.5 for a 2.5% Zusatzbeitrag. Bundesland must be the full Obolus state name such as Berlin. Do not send raw German Obolus flags such as gesetzliche_RV, gesetzlicheKvPvStatus, KV_Art, or KVZ in ChatGPT-facing calls; the adapter maps the semantic fields internally.

taxcompare

ChatGPT
Use this as the default tool whenever the user asks to compare salary, annual net income, tax burden, social contributions, or effective burden across multiple countries. Do not use berechne for multi-country comparisons unless the user asks for detailed country-specific payroll. If the user asks for one salary across multiple countries, always use taxcompare first. Use this endpoint for multi-country salary/tax comparisons. Select taxcompare when the user provides one salary or income amount, two or more countries, and wants tax, net salary, deductions, take-home pay, ranking, or 'where do I keep more money' results. Typical trigger phrases include: calculate tax for 50k in Germany, Ireland, USA, Canada; compare net salary across these countries; where do I pay less tax on EUR 80k; what is take-home pay in X, Y, Z; rank these countries by net income; how much do I keep after tax in Germany vs Ireland. Routing rule: if countries.length >= 2 and the same salary/income is used, use taxcompare. If countries.length === 1, use berechne. If the user provides detailed payroll assumptions such as tax class, province/state, health insurance, children, church tax, pension exemption, deductions, or benefits, use berechne unless taxcompare supports those fields explicitly. Do not use berechne repeatedly for simple multi-country comparisons unless taxcompare cannot accept the required inputs. Use shared_gross when the user wants to compare the same salary offer across countries. Prefer salary_ct in minor units, such as 5000000 for EUR 50,000. annual_gross is a backward-compatible annual major-unit alias, such as 50000 for EUR 50,000. Use local_median_gross when the user wants a local-market benchmark. In this mode Obolus fans out to country-specific editorial median gross benchmarks; annual_gross is optional and ignored when omitted. taxcompare is annual and country-level. It is not the right tool for detailed payroll inputs such as German KVZ, tax class IV factor, private health insurance, commuter allowance, or monthly gross; use berechne for that. countries accepts DE, AT, CH, US, UK, IE, CA, and AU. currency accepts EUR, USD, CHF, CAD, AUD, GBP and lower-case equivalents; results normalize currency codes to lower-case. joint_assessment and children are broad comparison assumptions. Country-specific filing details are simplified for comparability; for a precise single-country case, call berechne. Results are annual major-currency estimates. net is annual take-home pay; tax is annual tax burden; social_contributions are annual employee payroll/social deductions or the country-specific equivalent; effective_rate is a percentage-point burden value such as 37.7 for 37.7%. Country meanings differ: DE/AT emphasize income tax plus statutory social contributions, CH can include withholding and payroll social deductions but not private KVG/LAMal health bills, US includes federal/state and OASDI/Medicare-style items, UK includes PAYE/National Insurance, IE includes PAYE/USC/PRSI, CA includes federal/provincial tax plus CPP/QPP/EI, and AU includes PAYG/Medicare/super-related outputs.

Capabilities

Interactive

App Stats

2

Tools

ChatGPT

Platforms

Works with

ChatGPT

Data refreshed daily