API marketplace

Rocket-propulsion maths as an API, computed locally and deterministically. The delta-v endpoint applies the Tsiolkovsky rocket equation, Δv = ve·ln(m0/mf) with the exhaust velocity ve = Isp·g0, to give the velocity change a stage can produce from its wet (fuelled) mass, dry (burnout) mass and specific impulse — the delta-v budget that determines which manoeuvres are possible. The mass-ratio endpoint inverts the equation to give the mass ratio m0/mf = exp(Δv/ve) and the propellant mass fraction required to achieve a target delta-v, and, given a dry mass, the wet mass and propellant needed — revealing the steep, exponential tyranny of the rocket equation. The burn endpoint computes the propellant mass-flow rate ṁ = thrust/ve, the burn time and the total impulse from the thrust and propellant mass, and the delta-v if the wet mass is given. Masses are in kilograms, specific impulse in seconds, exhaust velocity and delta-v in metres per second and thrust in newtons, with standard gravity g0 = 9.80665 m/s². Everything is computed locally and deterministically, so it is instant and private. Ideal for aerospace, model-rocketry, spaceflight-simulation and orbital-mission app developers, stage-sizing and trajectory tools, and physics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is rocket propulsion; for orbital velocity and escape velocity use an orbital-mechanics API.

Uptime

100.0%

Latency

109ms

Subs

3,845

Building-acoustics soundproofing maths as an API, computed locally and deterministically. The mass-law endpoint computes the sound-transmission loss of a single partition from its surface mass density and the frequency using the field-incidence mass law, TL = 20·log10(m·f) − 47 dB — transmission loss rises about 6 dB for every doubling of mass or of frequency — and also gives the normal-incidence value. The composite endpoint combines the transmission losses of several elements that make up one wall, such as a heavy wall with a window or a door, by area-weighting their transmission coefficients, TL = −10·log10(Σ(Ai·τi)/ΣAi) — which shows how the weakest element, like a small gap or a thin window, dominates and wrecks an otherwise good wall. The transmission endpoint computes the received sound level on the far side of a partition, the source level minus the transmission loss, with an optional room-to-room correction that adds 10·log10(partition area / receiving-room absorption). Surface density is in kg/m², frequency in Hz, levels and transmission losses in dB and areas in m². Everything is computed locally and deterministically, so it is instant and private. Ideal for architecture, building-acoustics, studio-design, HVAC-noise and construction app developers, partition and noise-control tools, and acoustics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is sound insulation; for room reverberation use a reverberation API and for sound pressure level a sound-level API.

Uptime

100.0%

Latency

74ms

Subs

3,083

Transmission-line RF maths as an API, computed locally and deterministically for a lossless line. The input-impedance endpoint transforms a complex load impedance along a line, Zin = Z0·(ZL + jZ0·tanβl)/(Z0 + jZL·tanβl), from the characteristic impedance, the load resistance and reactance and the electrical length in degrees — a quarter-wave (90°) line inverts the load to Z0²/ZL while a half-wave (180°) line repeats it, which is the basis of impedance matching. The quarter-wave endpoint computes the characteristic impedance Z0 = √(Z1·Z2) of a quarter-wave transformer that matches two real impedances, exact at one frequency. The electrical-length endpoint converts a physical line length to its electrical length in wavelengths, degrees and radians at a frequency, using the on-line wavelength λ = vf·c/f with a velocity factor for the dielectric. Impedances are in ohms (the load split into resistance and reactance), electrical length in degrees, physical length in metres and frequency in hertz. Everything is computed locally and deterministically, so it is instant and private. Ideal for RF, antenna-matching, PCB, radar and microwave app developers, stub-matching and transformer-design tools, and electromagnetics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is line impedance transformation; for SWR and return loss use a VSWR API and for microstrip trace geometry a PCB API.

Uptime

100.0%

Latency

82ms

Subs

3,598

Rectangular-waveguide microwave maths as an API, computed locally and deterministically. The cutoff endpoint computes the cutoff frequency fc = (c/2)·√((m/a)²+(n/b)²) and cutoff wavelength of any TEmn or TMmn mode of a rectangular waveguide of inner width a and height b — below the cutoff a mode is evanescent and cannot propagate, and for the usual a > b the dominant mode is TE10 with fc = c/(2a). The guide-wavelength endpoint computes, at an operating frequency, the free-space wavelength, the guide wavelength λg = λ0/√(1−(fc/f)²) which is longer than free space, and the phase velocity (greater than c) and group velocity (the energy speed, below c). The modes endpoint lists every mode that propagates at a given frequency, sorted by cutoff, and identifies the dominant mode — so single-mode operation needs the frequency between the first and second cutoffs. Dimensions are in millimetres and frequencies in gigahertz, with c = 299,792,458 m/s. Everything is computed locally and deterministically, so it is instant and private. Ideal for RF, microwave, radar, satellite and antenna-feed app developers, waveguide-band and component-design tools, and electromagnetics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is metallic rectangular waveguide; for optical-fibre guiding use a fibre API and for SWR a VSWR API.

Uptime

100.0%

Latency

76ms

Subs

3,109

Optical-fibre photonics maths as an API, computed locally and deterministically. The numerical-aperture endpoint computes a step-index fibre's numerical aperture NA = √(n1² − n2²) from the core and cladding refractive indices, the acceptance angle θa = arcsin(NA) — the half-angle of the cone of light the fibre can capture — the full acceptance cone and the relative index difference Δ = (n1 − n2)/n1. The v-number endpoint computes the normalized frequency V = 2π·a·NA/λ from the core radius, the numerical aperture (or the indices) and the wavelength, classifies the fibre as single-mode when V is below the 2.405 cutoff or multimode above it, and gives the cutoff wavelength for single-mode operation. The modes endpoint estimates the number of guided modes — about V²/2 for a step-index fibre and V²/4 for a graded-index one — and confirms single-mode operation below the cutoff. Core radius and wavelength are in metres (1310 nm = 1.31×10⁻⁶ m) and refractive indices are dimensionless. Everything is computed locally and deterministically, so it is instant and private. Ideal for telecom, photonics, datacenter, sensor and laser app developers, fibre-link and waveguide-design tools, and optics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is optical-fibre guiding; for thin lenses and mirrors use a lens API and for refraction at a surface a Snell API.

Uptime

100.0%

Latency

81ms

Subs

3,629

Inductor-design electromagnetics as an API, computed locally and deterministically. The solenoid endpoint computes the inductance of a straight coil with the long-solenoid formula L = μ₀·μr·N²·A/l, from the number of turns, the coil length, the cross-sectional area (or diameter) and the relative permeability of the core — a ferromagnetic core multiplies the inductance. The toroid endpoint computes the inductance of a doughnut-shaped coil of rectangular cross-section, L = μ₀·μr·N²·h·ln(b/a)/(2π), from the turns, the axial height and the inner and outer radii; the toroidal shape confines the magnetic flux so there is little stray field. The energy endpoint computes the magnetic energy stored in an inductor, E = ½·L·I², and the flux linkage Φ = L·I, from the inductance and current — the energy released when the current is interrupted causes the inductive kick. Lengths are in metres, area in square metres, inductance in henries (millihenries and microhenries also returned) and current in amps, with μ₀ = 4π×10⁻⁷ H/m. Everything is computed locally and deterministically, so it is instant and private. Ideal for electronics, RF, power-supply, filter and motor-design app developers, coil-winding and inductor-sizing tools, and electromagnetics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is inductance from geometry; for the resonant frequency and reactance use a resonance API and for full AC impedance an impedance API.

Uptime

100.0%

Latency

77ms

Subs

3,325

Blast-effects and TNT-equivalence maths as an API for safety engineering and education, computed locally and deterministically. The energy endpoint converts between a TNT charge mass and the energy it releases using the conventional 4.184 MJ per kilogram, in both directions, including the kiloton equivalent. The scaled-distance endpoint computes the Hopkinson-Cranz scaled distance Z = R / W^(1/3) from a standoff distance and a charge weight — the cube-root scaling law means two blasts with the same Z produce the same overpressure — and inverts it to give the standoff distance for a target Z, which is how safety distances are set for a given charge. The overpressure endpoint estimates the peak side-on overpressure with the Brode (1955) correlation from the scaled distance (or from distance and charge), in kilopascals, bar and psi, with a qualitative damage assessment from shattered glass to structural collapse. Charge is in kilograms of TNT, distance in metres and energy in joules. Everything is computed locally and deterministically, so it is instant and private. Ideal for blast-resistant design, demolition, mining, process-safety and emergency-planning app developers, standoff-distance and overpressure tools, and engineering education; for engineering use, consult the applicable standards. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is explosive blast effects; for earthquake magnitude and energy use an earthquake-magnitude API.

Uptime

100.0%

Latency

78ms

Subs

3,051

Combinatorics maths as an API, computed locally and deterministically with exact arbitrary-precision integers. The factorial endpoint computes n! = 1·2·3···n (with 0! = 1) and returns it exactly as a string together with its digit count, so even very large factorials stay precise. The permutations endpoint counts ordered arrangements: without repetition nPr = n!/(n−r)! arrangements of r items chosen from n, and with repetition n^r, where each of the r positions may be any of the n items. The combinations endpoint counts unordered selections: without repetition the binomial coefficient nCr = n!/(r!·(n−r)!), and with repetition (multisets) C(n+r−1, r), where repeats are allowed. All results are computed with BigInt so they are exact no matter how large, returned as a string with the number of digits and a floating-point approximation when it fits. n and r are non-negative integers up to 100000. Everything is computed locally and deterministically, so it is instant and private. Ideal for probability, statistics, lottery, game-design, cryptography and education app developers, counting and odds tools, and discrete-maths teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is counting combinatorics; for modular arithmetic use a modular API and for descriptive statistics a statistics API.

Uptime

100.0%

Latency

82ms

Subs

4,241

Inflation-economics maths as an API, computed locally and deterministically. The adjust endpoint expresses a value across time in two ways — by an annual inflation rate over a number of years, V = amount·(1+r)^years, or by a ratio of consumer-price-index figures, V = amount·CPI_end/CPI_start — so an old price can be restated in today's money, with the total inflation over the period. The real-rate endpoint computes the real (inflation-adjusted) interest or investment rate from a nominal rate and an inflation rate using the Fisher equation, 1 + real = (1 + nominal)/(1 + inflation), alongside the rough nominal-minus-inflation approximation. The purchasing-power endpoint shows how inflation erodes money over time — the future buying power of today's amount, amount/(1+r)^years, the value lost and the larger amount needed to maintain the same purchasing power. Rates may be entered as a percent or a fraction and amounts in any currency. Everything is computed locally and deterministically, so it is instant and private. Ideal for personal-finance, budgeting, salary, retirement-planning and economics app developers, cost-of-living and real-return tools, and finance education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is inflation adjustment; for loan repayments use a loan API and for investment growth an investment API.

Uptime

100.0%

Latency

69ms

Subs

3,900

Colligative-properties chemistry maths as an API, computed locally and deterministically. The freezing-point endpoint computes the freezing-point depression ΔTf = i·Kf·m and the resulting lowered freezing point of a solution, from the molality, the cryoscopic constant (1.86 °C·kg/mol for water) and the van 't Hoff factor i — which is 1 for a non-electrolyte like sugar, about 2 for sodium chloride and about 3 for calcium chloride. The boiling-point endpoint computes the boiling-point elevation ΔTb = i·Kb·m and the raised boiling point, with the ebullioscopic constant (0.512 °C·kg/mol for water). The osmotic-pressure endpoint computes the van 't Hoff osmotic pressure Π = i·M·R·T from the molarity, the temperature and the van 't Hoff factor, the pressure that drives osmosis across a semipermeable membrane, returned in atmospheres, kilopascals and bar. Molality is in mol per kg of solvent, molarity in mol per litre of solution and temperature in kelvin. Everything is computed locally and deterministically, so it is instant and private. Ideal for chemistry-education, food-science, antifreeze, desalination and biology app developers, solution and de-icing tools, and STEM teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is colligative properties of solutions; for a compound's molar mass use a molar-mass API and for dilution concentrations a dilution API.

Uptime

100.0%

Latency

80ms

Subs

4,216

Morse code conversion as an API, computed locally and deterministically. The encode endpoint turns text into International Morse code, mapping A–Z, the digits 0–9 and common punctuation to dots and dashes, separating letters with a space and words with a slash, and listing any unsupported characters it skipped. The decode endpoint turns Morse code back into text, accepting word separators written as a slash, a pipe or a wide gap, and marking unrecognised symbols. The timing endpoint computes the PARIS-standard timing from a words-per-minute speed — the dot duration is 1200/WPM milliseconds, a dash is three dots, and the gaps are one, three and seven dot units for intra-character, inter-character and word spacing — and, given a Morse message, the total number of units and the transmission time. The word PARIS is exactly 50 units, which defines the WPM scale. Everything is computed locally and deterministically, so it is instant and private. Ideal for amateur-radio, aviation, education, accessibility, puzzle and game app developers, signalling and CW-training tools, and learning Morse. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is Morse code; for Base64 and JWT use an encoding API and for Caesar and substitution ciphers a cipher API.

Uptime

100.0%

Latency

84ms

Subs

3,080

Mechanical-fatigue engineering maths as an API, computed locally and deterministically. The stress-cycle endpoint decomposes a cyclic load given by its maximum and minimum stress into the alternating stress σa = (σmax − σmin)/2, the mean stress σm = (σmax + σmin)/2, the stress range and the stress ratio R = σmin/σmax, and names the loading (fully reversed at R = −1, repeated at R = 0). The criteria endpoint computes the infinite-life safety factor against fatigue using the three classic mean-stress theories — Goodman (1/n = σa/Se + σm/Sut, standard and safe), Soderberg (uses the yield strength, conservative) and Gerber (a parabola, least conservative) — from the alternating and mean stress, the endurance limit Se, the ultimate strength Sut and an optional yield strength. The endurance-limit endpoint estimates the corrected endurance limit Se = ka·kb·kc·kd·ke·Se' from the ultimate strength, with Se' = 0.5·Sut for steel and the Marin modifying factors for surface finish, size, load type, temperature and reliability. Stresses and strengths use any one consistent unit (MPa is typical). Everything is computed locally and deterministically, so it is instant and private. Ideal for mechanical, structural, automotive and aerospace-design app developers, durability and safety-factor tools, and engineering education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is fatigue and endurance; for static stress transformation use a Mohr-circle API and for column buckling a buckling API.

Uptime

100.0%

Latency

71ms

Subs

4,394

Hydroelectric-power engineering maths as an API, computed locally and deterministically. The power endpoint computes the electrical power a hydro plant generates with P = ρ·g·Q·H·η, from the water flow rate, the net head (the effective drop), the overall turbine-generator efficiency (typically 0.80–0.92) and the water density, returning both the gross power at 100 % efficiency and the net electrical output. The sizing endpoint inverts the relation to size a scheme — given a target power it solves the flow rate needed at a known head, or the head needed at a known flow, Q = P/(ρ·g·H·η). The annual-energy endpoint computes the yearly energy from the rated power and a capacity factor (typically 0.3–0.6 for hydro, accounting for water availability and downtime), E = P × 8760 h × capacity factor, and an optional revenue from an electricity price. Flow is in cubic metres per second, head in metres, efficiency 0–1, power in watts, kilowatts and megawatts. Everything is computed locally and deterministically, so it is instant and private. Ideal for renewable-energy, micro-hydro, civil-engineering, feasibility and sustainability app developers, run-of-river and reservoir tools, and energy education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is hydroelectric generation; for wind-turbine power use a wind-power API, for solar resource a solar API and for pump (energy-consuming) duty a pump API.

Uptime

100.0%

Latency

85ms

Subs

3,555

Roman numeral conversion as an API, computed locally and deterministically. The encode endpoint turns an integer from 1 to 3999 into its Roman numeral using standard subtractive notation, so 1994 becomes MCMXCIV and 2024 becomes MMXXIV. The decode endpoint turns a Roman numeral back into an integer with strict validation — it rejects malformed forms such as IIII or VV and also returns the canonical way to write the same value, accepting any letter case. The arithmetic endpoint adds, subtracts or multiplies two values given as either integers or Roman numerals and returns the result as a Roman numeral and as an integer, provided the result stays within the classic 1–3999 range. The standard subtractive pairs are IV, IX, XL, XC, CD and CM. Everything is computed locally and deterministically, so it is instant and private. Ideal for typesetting, publishing, education, clock-face, game and document-processing app developers, numbering and chapter tools, and history teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is Roman numeral conversion; for binary, octal and hexadecimal number-base conversion use a base-conversion API.

Uptime

100.0%

Latency

77ms

Subs

3,999

The Beaufort wind scale as an API, computed locally and deterministically. The classify endpoint turns a measured wind speed — in metres per second, kilometres per hour, knots, miles per hour or feet per second — into its Beaufort force (0 calm to 12 hurricane), with the descriptive name (light breeze, gale, storm …), the corresponding sea state and the mean open-sea wave height, plus the speed expressed in every unit. The force endpoint looks up a Beaufort number and returns its wind-speed range in all units, its description, sea condition and wave height. The convert endpoint converts a wind speed across metres per second, kilometres per hour, knots, miles per hour and feet per second and reports the matching Beaufort force (1 knot = 0.514444 m/s). Speeds use the standard 10-metre reference height and wave heights are open-sea means. Everything is computed locally and deterministically, so it is instant and private. Ideal for sailing, marine, aviation, drone, weather and outdoor app developers, wind-warning and sea-state tools, and meteorology education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is the Beaufort wind scale; for the feels-like wind chill use a feels-like API and for live wind observations a weather data API.

Uptime

100.0%

Latency

79ms

Subs

3,279

Feels-like (apparent) temperature meteorology as an API, computed locally and deterministically. The wind-chill endpoint computes how cold the air feels when wind carries body heat away, using the Environment Canada formula WC = 13.12 + 0.6215·T − 11.37·V^0.16 + 0.3965·T·V^0.16 from the air temperature (°C) and wind speed (km/h), valid at 10 °C or below with wind of at least 4.8 km/h. The heat-index endpoint computes how hot it feels in warm, humid air with the US National Weather Service Rothfusz regression from temperature and relative humidity, since high humidity slows sweat evaporation, with the low-/high-humidity adjustments. The apparent-temperature endpoint computes the Australian Bureau of Meteorology apparent temperature, AT = Ta + 0.33·e − 0.70·ws − 4.00, which combines the warming effect of humidity (through the vapour pressure e) and the cooling effect of wind (ws in m/s) in a single feels-like value. Temperatures are in °C (Fahrenheit also returned), humidity in %, wind in km/h for wind chill and m/s for apparent temperature. Everything is computed locally and deterministically, so it is instant and private. Ideal for weather, outdoor-activity, sports, smart-home and wearable app developers, comfort and safety tools, and meteorology education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is the feels-like temperature calculator; for the occupational WBGT heat-stress index use a WBGT API and for live weather observations a weather data API.

Uptime

100.0%

Latency

72ms

Subs

4,278

Earthquake-magnitude seismology as an API, computed locally and deterministically. The energy endpoint computes the radiated seismic energy released by an earthquake of a given magnitude using the Gutenberg-Richter relation, log10(E) = 1.5·M + 4.8 with E in joules, and converts it to a TNT equivalent in tons and kilotons (one ton of TNT ≈ 4.184×10⁹ J), with a felt/damage classification. The compare endpoint quantifies how much bigger one quake is than another: each magnitude unit means about ten times the ground-motion amplitude on a seismograph and about 31.6 times (10^1.5) the energy, so it returns both the amplitude ratio and the energy ratio between two magnitudes. The moment-magnitude endpoint converts between the seismic moment M0 (in newton-metres, M0 = rigidity × rupture area × slip) and the moment magnitude with the Hanks-Kanamori relation Mw = (2/3)·log10(M0) − 6.07, in either direction. Magnitudes are dimensionless, energy is in joules and seismic moment in newton-metres. Everything is computed locally and deterministically, so it is instant and private. Ideal for seismology-education, disaster-modelling, insurance, structural-risk and science app developers, earthquake-energy and magnitude tools, and STEM teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is the earthquake-magnitude calculator; for real-time and historical earthquake event feeds use an earthquake data API.

Uptime

100.0%

Latency

74ms

Subs

4,970

Helmholtz-resonator acoustics as an API, computed locally and deterministically. The frequency endpoint computes the resonant frequency of a Helmholtz resonator — a cavity with a neck, like a bottle or a ported speaker box — from the neck area (or diameter), the neck length and the cavity volume, f = (c/2π)·√(A/(V·L_eff)), adding the acoustic end correction (about 0.85·radius for a flanged end and 0.61·radius for a free end) so a short or open neck resonates lower than its physical length suggests. The design endpoint inverts the relation, V = A·c²/(L_eff·ω²), to give the cavity volume needed to tune a resonator or a muffler chamber to a target frequency. The port-tuning endpoint sizes a bass-reflex (vented loudspeaker) box port in practical audio units — from the box volume in litres and the port diameter in centimetres it gives the tuning frequency for a given port length, or the port length required for a target tuning frequency, using the 0.732·diameter end correction. Core endpoints use SI units; the speed of sound defaults to 343 m/s. Everything is computed locally and deterministically, so it is instant and private. Ideal for audio, loudspeaker-design, musical-instrument, muffler and acoustic-treatment app developers, bass-reflex and resonator tools, and acoustics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is Helmholtz resonance; for room reverberation use a reverberation API and for standing waves on strings and in pipes a standing-wave API.

Uptime

100.0%

Latency

78ms

Subs

3,171

Solar-position astronomy as an API, computed locally and deterministically with the NOAA solar-calculator algorithm. The position endpoint gives the sun's elevation (altitude above the horizon), azimuth (clockwise from true north), zenith angle and hour angle for any latitude, longitude, date and local time with a UTC offset — telling you exactly where the sun is in the sky and whether it is above the horizon. The declination endpoint gives the solar declination — the sun's angle north or south of the equator, about +23.44° at the June solstice and −23.44° in December — and the equation of time, the difference between apparent and mean solar time, for any date. The solar-noon endpoint gives the local clock time of solar noon, the peak (noon) elevation 90 − |latitude − declination| and the day length, handling polar day and polar night. Latitudes and longitudes are in degrees (north and east positive), dates are YYYY-MM-DD and times HH:MM:SS local. Everything is computed locally and deterministically, so it is instant and private. Ideal for solar-tracking, PV-panel-orientation, photography golden-hour, agriculture, shading-analysis and astronomy app developers, sun-path and daylight tools, and STEM teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is the sun's position in the sky; for sunrise and sunset clock times use a sunrise API and for solar irradiance and PV resource a solar-resource API.

Uptime

100.0%

Latency

75ms

Subs

4,086

Load-cell (weighing-transducer) maths as an API, computed locally and deterministically. The output endpoint computes the bridge output voltage a strain-gauge load cell produces under a given load, Vout = (load/capacity)·sensitivity·excitation, where the full-scale output FSO = sensitivity(mV/V)·excitation(V) is reached at the rated capacity — it returns the output in millivolts, the equivalent mV/V at that load and the capacity utilization, and flags overload. The load endpoint inverts this to recover the applied load from a measured bridge output, load = (Vout/FSO)·capacity. The array endpoint sizes a multi-cell weighing platform: from the number of identical cells, the per-cell capacity and the live and dead (tare) load it returns the evenly distributed per-cell load, its output and utilization and the total system capacity, so cells can be chosen to stay under capacity in the worst case. Sensitivity is in mV/V, excitation in volts (default 10), output in millivolts; load and capacity share any consistent unit. Everything is computed locally and deterministically, so it is instant and private. Ideal for industrial-weighing, scale, force-measurement, silo and process-control app developers, load-cell sizing and calibration tools, and instrumentation education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is load-cell transducer output; for the underlying Wheatstone-bridge and strain maths use a Wheatstone-bridge API.

Uptime

100.0%

Latency

72ms

Subs

4,972

AC complex-impedance maths as an API, computed locally and deterministically. The series endpoint computes the impedance of a series R-L-C circuit at a given frequency — the inductive reactance X_L = 2πf·L, the capacitive reactance X_C = 1/(2πf·C), the complex impedance Z = R + j(X_L − X_C), its magnitude |Z| = √(R²+X²) and phase angle φ = atan(X/R) — and classifies the circuit as inductive (current lags), capacitive (current leads) or resistive. The parallel endpoint computes a parallel R-L-C impedance through its admittance Y = 1/R + j(ωC − 1/ωL) and Z = 1/Y, with magnitude and phase. The ac-ohm endpoint applies Ohm's law for AC, I = V / |Z|, to give the RMS current and apparent power from an RMS voltage and an impedance specified either as resistance and reactance or as a magnitude, and the real power when the phase is known. Resistance and reactance are in ohms, inductance in henries, capacitance in farads, frequency in hertz and voltage RMS in volts; phase is in degrees. Everything is computed locally and deterministically, so it is instant and private. Ideal for electronics, audio, RF-filter, power-supply and motor-control app developers, AC-circuit and phasor tools, and electrical-engineering education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is AC complex impedance; for the resonant frequency and reactance alone use a resonance API and for power-factor correction a power-factor API.

Uptime

100.0%

Latency

90ms

Subs

4,703

Uniform circular-motion physics as an API, computed locally and deterministically. The centripetal-force endpoint computes the centripetal acceleration a = v²/r = ω²·r — always pointing toward the centre — and the centripetal force F = m·a that holds a body on its circular path, from the mass, the radius and either the linear or the angular velocity, and reports the equivalent g-force. The angular endpoint converts between every way of describing rotation — angular velocity (rad/s), revolutions per minute, frequency, period and, given a radius, the linear (tangential) velocity — using ω = 2π·f = 2π/T = v/r. The centrifuge endpoint computes the relative centrifugal force (RCF, in g) of a centrifuge rotor from its speed in rpm and radius, RCF = ω²·r / g, or inverts it to give the rpm needed to reach a target RCF. Masses are in kg, radii in m (mm for the centrifuge), velocities in m/s, angular velocities in rad/s and forces in N. Everything is computed locally and deterministically, so it is instant and private. Ideal for physics-education, mechanical, automotive, lab-centrifuge and amusement-ride app developers, rotational-motion and g-force tools, and STEM teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is uniform circular motion; for gravitational orbits use a gravitation API, for a vehicle on a banked curve a banked-curve API and for pendulum oscillation a pendulum API.

Uptime

100.0%

Latency

74ms

Subs

3,462

NTC-thermistor sensor maths as an API, computed locally and deterministically. The steinhart-hart endpoint converts between resistance and temperature using the Steinhart-Hart equation, 1/T = A + B·ln R + C·(ln R)³ — the most accurate NTC model — in both directions, solving the resistance at a given temperature with Cardano's cubic formula. The beta endpoint uses the simpler two-point Beta model, 1/T = 1/T0 + (1/β)·ln(R/R0) and R = R0·exp(β·(1/T − 1/T0)), to convert resistance to temperature or back from a reference resistance R0 at T0 (default 25 °C) and the beta coefficient. The divider endpoint recovers the thermistor's resistance from a voltage-divider reading — low-side R = Rs·Vout/(Vsupply − Vout) or high-side — so an ADC voltage can be turned into a resistance and then a temperature. Resistance is in ohms, temperature in °C (kelvin also returned), voltages in volts and beta in kelvin. Everything is computed locally and deterministically, so it is instant and private. Ideal for embedded, IoT, HVAC-control, 3D-printer and battery-management app developers, temperature-sensing and calibration tools, and electronics education. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is NTC thermistor conversion; for a generic resistive divider use an LED-resistor or voltage-drop API and for thermal expansion a thermal-expansion API.

Uptime

100.0%

Latency

77ms

Subs

4,467

Chemical reaction-stoichiometry maths as an API, computed locally and deterministically. The limiting-reagent endpoint takes two reactants with their amounts in moles and their balanced-equation coefficients and finds which one runs out first — the limiting reagent — by comparing the moles/coefficient ratio (the reaction extent), and returns how much of the excess reagent is left over. The yield endpoint computes the theoretical yield of a product, in moles and grams, from the limiting reagent and the product's stoichiometric coefficient and molar mass, n_product = n_limiting·(coeff_product/coeff_limiting), and — given the actual yield — the percent yield. The mole-mass endpoint converts between moles, mass and the number of particles for a given molar mass, using moles = mass / molar_mass and N = moles · Avogadro's number (6.02214076e23). Amounts are in moles, masses in grams and molar masses in g/mol. Everything is computed locally and deterministically, so it is instant and private. Ideal for chemistry-education, lab, pharmaceutical and chemical-engineering app developers, reaction-planning and yield tools, and STEM teaching. Pure local computation — no key, no third-party service, instant. Live, nothing stored. 3 endpoints. This is reaction stoichiometry; for a compound's molar mass from its formula use a molar-mass API and for solution concentrations a dilution API.

Uptime

100.0%

Latency

76ms

Subs

4,333