fairchem-core
Machine learning models for chemistry and materials science by the FAIR Chemistry team
Decision gist · record as of 2026-08-14
Yes, if you work in computational chemistry or materials science and want fast, pretrained ML-based property predictions integrated with ASE. Requires GPU access, a Hugging Face account, and familiarity with the ASE ecosystem. Active maintenance, permissive license, and no known vulnerabilities support adoption. Version 2 is a breaking change—verify compatibility with your existing workflows before upgrading from version 1.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires CUDA-capable GPU and Hugging Face account with access to the UMA model repository; login via huggingface-cli login before use.
- Low friction installation with a pure Python wheel.
- Active maintenance with recent releases; last commit 2026-08-14.
License · maintenance · safety
MIT License (permissive) — MIT License permits commercial use, modification, and redistribution with minimal restrictions. Suitable for both research and production applications.
last release 2026-06-08 (67 days) · last repo commit 2026-08-14 · 2,220 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 110,269 downloads/mo, #12,474 on PyPI
Alternatives
Verify before relying
pip install fairchem-core
from fairchem.core import pretrained_mlip, FAIRChemCalculator
from ase.build import bulk
from ase.optimize import FIRE
predictor = pretrained_mlip.get_predict_unit("uma-s-1p2", device="cuda")
calc = FAIRChemCalculator(predictor, task_name="omat")
atoms = bulk("Fe")
atoms.calc = calc
opt = FIRE(atoms)
opt.run(0.05, 100)- Whether multi-GPU inference (workers=N flag) requires additional configuration beyond Ray installation.
- Compatibility of UMA-1.2 models with existing downstream workflows and whether version 2 breaking changes affect common use patterns.
- Performance characteristics and memory requirements for different model sizes (uma-s vs uma-m) on typical hardware.
What it is and what it does
Fairchem-core is FAIR Chemistry's centralized repository of machine learning models for materials science and quantum chemistry. It provides pretrained models (UMA series) that can predict molecular and materials properties across multiple domains: catalysis, oxides, molecules, polymers, inorganic crystals, MOFs, and molecular crystals. The package integrates with ASE (Atomic Simulation Environment) via the FAIRChemCalculator, allowing users to perform structure relaxations, molecular dynamics simulations, and property calculations by selecting a task name appropriate to their domain.
The package depends on a large scientific stack including torch, ray, scipy, and specialized libraries like e3nn and hydra-core. It supports multi-GPU and multi-node inference for large-scale simulations. Version 2 is a breaking change from version 1 and incompatible with older pretrained models. The latest UMA-1.2 release (March 2026) claims approximately 50% faster inference and expanded data coverage for catalysts, molecules, and polymers.
Use it for
- Relax adsorbate-surface systems for heterogeneous catalysis research without running expensive DFT calculations.
- Perform molecular dynamics simulations of molecules and polymers at scale using GPU acceleration.
- Predict formation energies and structural stability of inorganic crystals for materials discovery workflows.
- Calculate spin gaps and other quantum properties of molecules for computational chemistry studies.
- Run large-scale MD simulations (100k+ atoms) on multi-GPU clusters for industrial materials modeling.
- Estimate properties of metal-organic frameworks and molecular crystals for screening applications.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you work in computational chemistry or materials science and want fast, pretrained ML-based property predictions integrated with ASE.
Requires GPU access, a Hugging Face account, and familiarity with the ASE ecosystem. Active maintenance, permissive license, and no known vulnerabilities support adoption. Version 2 is a breaking change—verify compatibility with your existing workflows before upgrading from version 1.
Install
fairchem-core on PyPI
Before you install
Low friction installation with a pure Python wheel. Active maintenance with recent releases; last commit 2026-08-14. Requires 22 runtime dependencies including torch, ray, and scientific computing libraries—a substantial but standard stack for ML-driven materials science.
Requires CUDA-capable GPU and Hugging Face account with access to the UMA model repository; login via huggingface-cli login before use.
License in practice
MIT License permits commercial use, modification, and redistribution with minimal restrictions. Suitable for both research and production applications.
Quickstart
pip install fairchem-core
from fairchem.core import pretrained_mlip, FAIRChemCalculator
from ase.build import bulk
from ase.optimize import FIRE
predictor = pretrained_mlip.get_predict_unit("uma-s-1p2", device="cuda")
calc = FAIRChemCalculator(predictor, task_name="omat")
atoms = bulk("Fe")
atoms.calc = calc
opt = FIRE(atoms)
opt.run(0.05, 100)Verify before relying
- Whether multi-GPU inference (workers=N flag) requires additional configuration beyond Ray installation.
- Compatibility of UMA-1.2 models with existing downstream workflows and whether version 2 breaking changes affect common use patterns.
- Performance characteristics and memory requirements for different model sizes (uma-s vs uma-m) on typical hardware.
Package facts
| License | MIT License permissive |
| Python support | Supports the current Python release <3.14,>=3.11 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 22 packagesase-db-backendsaseclusterscopee3nnhuggingface-hubhydra-corelmdbmontynumbanumpyorjsonpyyamlrayrequestsscipysetuptoolssubmitittorchtnttorchtqdmwandbwebsockets |
| Maintenance | Actively maintained 67 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 110,269 / month, #12,474 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: fairchem_core-2.21.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “molecular dynamics calculator”
- fairchem-coreProvides pretrained machine learning models for predicting molecular…
- aseASE provides Python tools for setting up, running, and analyzing…
- mace-torchMACE trains and evaluates machine learning interatomic potentials…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also ase · chgnet · mace-torch · matscipy · pyocse · mp-api · dpdata · ase-db-backends · dscribe · OpenMM