Aurigene-AI/ChemFM-1B

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.1BQuant:BF16Context Size:2kPublished:Sep 10, 2026License:mitArchitecture:Transformer0.0K Open Weights Featherless Exclusive Cold

ChemFM-1B is a 1.1 billion parameter chemistry foundation model developed by ChemFM and mirrored by Aurigene AI. Trained on UniChem, this model is specifically designed for chemistry-related tasks. Its primary strength lies in applications such as molecule generation, property prediction, and reaction tasks, making it suitable for hit generation and lead optimization in drug discovery.

Loading preview...

Overview

ChemFM-1B is a 1.1 billion parameter chemistry foundation model, originally developed by ChemFM and mirrored by Aurigene AI. It is specifically trained on the UniChem dataset, making it highly specialized for various chemical applications. This model is intended for use in the discovery stage of drug development, particularly for hit generation and lead optimization.

Key Capabilities

  • Molecule Generation: Capable of generating novel molecular structures.
  • Property Prediction: Can predict various properties of chemical compounds.
  • Reaction Tasks: Suitable for tasks involving chemical reactions.

Use Cases

ChemFM-1B is particularly well-suited for:

  • Accelerating drug discovery processes.
  • Assisting in hit generation and lead optimization phases.
  • Research and development in computational chemistry.

For comprehensive details, including training data and citation information, users are directed to the upstream model card.

Finetunes
1 models