# Scimlate > SciMLate is an AI scientific computing platform that runs CarnotAI models for engineering and scientific workflows. ## Links - [Scimlate app](https://dev.scimlate.com): Run scientific ML models from a chat interface. - [Pricing](https://dev.scimlate.com/pricing): Plans, credits, and limits. - [Contact](https://dev.scimlate.com/contact): Support and sales enquiries. ## Models - 2DGC FP FWD: Predicts freezing point (T_freeze, kelvin) for hydrocarbon mixtures from a 2D-GC composition matrix (30 carbon numbers × 8 compound classes; 189 trained bins). Uses the physics-residual ONNX surrogate (`2dgc_freezing_point_fwd.onnx`). - 2DGC Dist JF FWD: Predicts ASTM D86 distillation curve temperatures (T05–T95, kelvin) for jet-fuel / kerosene-like mixtures from a 2D-GC composition matrix (30 carbon numbers × 13 compound classes). Uses the trained ONNX surrogate (`2dgc_distillation_jet_fuel_fwd.onnx`). - Branching Index: Computes the branching index (BI) for one or more hydrocarbons from SMILES strings and mole fractions. Supports a single pure component, multiple pure components in one run (each BI reported separately), or a true blend (mole-fraction-weighted BI). BI quantifies molecular… - Cetane Number: Predicts cetane number (CN) for hydrocarbon jet fuels from a two-dimensional gas chromatography (2D-GC) carbon-number distribution. The input is a matrix of class-resolved abundances by carbon number (C5-C24). A trained post-Brix ONNX neural network maps the 160-element feature… - ignition-delay-time-ch4-0d - ignition-delay-time-ch4-0d-inverse - IDT TPRF: Predicts ignition delay time (in milliseconds) for TPRF (Toluene Primary Reference Fuel) surrogate mixtures based on molecular structure (SMILES), composition, pressure, temperature, and equivalence ratio. Uses a trained ONNX neural network model. - Liquid ρ(T): Evaluates liquid-phase density (kg/m³) as a function of temperature (K) using Cheetah ``density_correlation`` data (Landolt-Bornstein / Wilhoit polynomial and Guggenheim forms). Accepts a pure-component SMILES string. Comma-separated components run several pure fuels in one… - 0D IDT CH4: Predicts 0D homogeneous ignition delay time (milliseconds) for methane (CH4) / air mixtures from pressure, temperature, equivalence ratio, and exhaust-gas recirculation (EGR). Uses the trained SIDT ONNX surrogate (`0d_idt_ch4_fwd.onnx`). - 0D IDT CH4 INV: Inverts 0D homogeneous methane (CH4) / air ignition delay. Given a target ignition delay (milliseconds) and three of {pressure, temperature, equivalence ratio, EGR}, solves for the missing condition by 1-D root-finding on the SIDT ONNX surrogate (`0d_idt_ch4_inv.onnx`). The… ## FAQ - **What is Scimlate?** An AI scientific computing platform for engineering and scientific workflows. - **Which scientific models does Scimlate expose?** Active registered models include: 2DGC FP FWD, 2DGC Dist JF FWD, Branching Index, Cetane Number, ignition-delay-time-ch4-0d, ignition-delay-time-ch4-0d-inverse, IDT TPRF, Liquid ρ(T), 0D IDT CH4, 0D IDT CH4 INV. - **What does 2DGC FP FWD do?** Predicts freezing point (T_freeze, kelvin) for hydrocarbon mixtures from a 2D-GC composition matrix (30 carbon numbers × 8 compound classes; 189 trained bins). Uses the physics-residual ONNX surrogate (`2dgc_freezing_point_fwd.onnx`).