Physics AI dataset discovery

Find ML-ready physics simulation datasets.

SimHunt is a curated discovery hub for public CFD, FEA, FSI, thermal, and multiphysics datasets ranked for physics AI and engineering foundation-model training.

For ML engineers

Find simulation datasets with task, solver, file, and readiness signals before committing time to data wrangling.

For CAE teams

Discover public CFD, FEA, and multiphysics data through a compact readiness layer.

For researchers

Locate sources for neural operators, surrogate models, field prediction, and engineering foundation-model experiments.

Top public previews

View search

Example OpenFOAM CFD benchmark dataset

Sample public preview for an ML-ready CFD dataset. Direct source URL is locked for verified subscribers.

Gold
CFDexternal_aerodynamics OpenFOAM
Source Zenodo
Readiness High
Access Source URL locked Verified subscribers only

arungovindneelan/openfoam-Agent-Dataset

foam-openfoam-agent We’re on a journey to advance and democratize artificial intelligence through open source and open science. arungovindneelan/openfoam-Agent-Dataset · Datasets at Hugging Face Hugging Face Models Datas...

Silver
CFD OpenFOAM
Source Hugging Face
Readiness Medium
Access Source URL locked Verified subscribers only

sandialabs/WEC-OpenFOAM-Open-Datasets

Contribute to sandialabs/WEC-OpenFOAM-Open-Datasets development by creating an account on GitHub. GitHub - sandialabs/WEC-OpenFOAM-Open-Datasets · GitHub Skip to content Navigation Menu Toggle navigation Sign in Appearan...

Silver
CFDFEM OpenFOAM
Source GitHub
Readiness Medium
Access Source URL locked Verified subscribers only

fluentassertions/fluentassertions.datasets

Extensions for Fluent Assertions for working with DataSet, DataTable and DataRow

Silver
CFD Fluent
Source GitHub
Readiness Medium
Access Source URL locked Verified subscribers only

High-quality simulation data is the bottleneck for physics AI

SimHunt helps researchers, ML engineers, and data scientists discover public engineering-simulation datasets, inspect readiness signals, and find trusted source platforms for training surrogate models, neural operators, and physics foundation models.