MATCHA — Matching Text via Contrastive Semantic Alignment

A learned text similarity metric that captures both semantic alignment and contradiction through contrastive training.

Paper: MATCHA | Model: HuggingFace | Code: GitHub

Installation

pip install matcha-metric

Usage

from matcha_metric import MATCHA

model = MATCHA.from_pretrained("Siran-Li/MATCHA")

# Score a pair of texts
similarity = model.score("A dog is playing in the park.", "A puppy is running outside.")
print(f"Similarity: {similarity:.4f}")

# Batch scoring
scores = model.score(
    ["The cat sat on the mat.", "It is raining outside."],
    ["A feline rested on the rug.", "The weather is sunny and clear."],
)

# Get embeddings directly
embeddings = model.encode(["Hello world", "Hi there"])

Citation

@inproceedings{li2026matcha,
  title={MATCHA: Matching Text via Contrastive Semantic Alignment},
  author={Li, Siran and Etoglu, Ece Sena and Eickhoff, Carsten and Bahrainian, Seyed Ali},
  booktitle={Findings of the Association for Computational Linguistics: ACL 2026},
  pages={21001--21018},
  year={2026}
}