ModernBERT Similarity Demo

Compute the similarity between two texts using ModernBERT. Choose between different pooling strategies for embedding generation.

Pooling Method

Mean Pooling: Averages all token embeddings (better for semantic similarity)
CLS Pooling: Uses only the [CLS] token embedding (faster, might miss context)

Model

Choose between the variants of ModernBERT
Might take a few seconds to load the model

Examples
Text 1 Text 2 Pooling Method Model