ModelLLMsvv1

LegalBERT

by Ilias Chalkidis et al. (Academic) · free · Last verified 2026-03-17

LegalBERT is a family of BERT models pre-trained on a diverse corpus of English legal texts, including legislation, court cases, and contracts. This specialized training allows it to significantly outperform general-purpose BERT models on downstream legal NLP tasks, establishing it as a foundational baseline for legal AI research and applications.

https://huggingface.co/nlpaueb/legal-bert-base-uncased ↗

B—Above Average

Adoption: BQuality: AFreshness: CCitations: AEngagement: F

Specifications

License: CC BY-SA 4.0
Pricing: free
Capabilities: legal-text-classification, named-entity-recognition-for-legal-documents, contract-clause-extraction, semantic-search-for-case-law, legal-question-answering, summarization-of-legal-texts, e-discovery-document-prioritization
Integrations: hugging-face-transformers, pytorch, tensorflow, spacy
Use Cases: [object Object], [object Object], [object Object], [object Object], [object Object]
API Available: Yes
Parameters: ~110M
Context Window: 512 tokens
Modalities: text
Training Cutoff: 2020
Tags: legal-tech, bert, transformer-model, nlp, foundational-model, document-analysis, contract-analysis, case-law, e-discovery, text-classification
Added: 2026-03-17
Completeness: 0.95%

Index Score

62.5

Adoption

Quality

Freshness

Citations

Engagement

Need help choosing the right model?

Get Expert Guidance

Explore the full AI ecosystem on Agents as a Service