MatBench
Materials property prediction benchmark suite (used for some chemistry-adjacent ML).
Composite score: 83.3
Rubric (1–5 per criterion)
rigor
5/5
coverage
3/5
maintenance
4/5
adoption
4/5
quality
5/5
accessibility
5/5
industry_relevance
3/5
Metadata
Stages
Developmental Candidate
Modalities
cross-modality
Task types
materials-property
License
MIT
First release
2020
Last updated
2025-02
Flags
none
Size & scope
- tasks: 13
- samples: 132000
Primary paper
Title
Benchmarking materials property prediction methods: the Matbench test set and Automatminer reference algorithm
Authors
Dunn A, Wang Q, Ganose A, Dopp D, Jain A
Year
2020
DOI / arXiv
Citations
380
Links
- Official site: https://matbench.materialsproject.org/
- GitHub: https://github.com/materialsproject/matbench
- Leaderboard: https://matbench.materialsproject.org/Leaderboards%20Per-Task/
Hosted by (initiatives)
Experts (primary authors / maintainers)
Groups (host labs / companies / consortia)
Related benchmarks
- none
Notes (honest caveats)
Materials-science benchmark; relevant for formulation / co-crystal work.