gliner2-multi-v1

<div align="center"> <a href="https://agent.fastino.ai" target="_blank" rel="noopener noreferrer"> <img src="image/GLiNER-2-Multi-HF-Banner.png" alt="Fastino AI - Fine-tune GLiNER with a single prompt" width="100%"/> </a> </div>

GLiNER2: Unified Schema-Based Information Extraction and Text Classification

Extract entities, classify text, parse structured data, and extract relations—all in one efficient model.

GLiNER2 unifies Named Entity Recognition, Text Classification, Structured Data Extraction, and Relation Extraction into a single 205M parameter model. It provides efficient CPU-based inference without requiring complex pipelines or external API dependencies.

Fine-tune via Fastino. Join discussions on Discord and Reddit.

✨ Why GLiNER2?

• 🎯 One Model, Four Tasks: Entities, classification, structured data, and relations in a single forward pass
• 💻 CPU First: Lightning-fast inference on standard hardware—no GPU required
• 🛡️ Privacy: 100% local processing, zero external dependencies

Installation

pip install gliner2

Usage

Entity Extraction

from gliner2 import GLiNER2

# Load the model
extractor = GLiNER2.from_pretrained("fastino/gliner2-multi-v1")

# Extract entities
text = "Apple CEO Tim Cook announced iPhone 15 in Cupertino yesterday."
result = extractor.extract_entities(text, ["company", "person", "product", "location"])

print(result)
# Output: {'entities': {'company': ['Apple'], 'person': ['Tim Cook'], 'product': ['iPhone 15'], 'location': ['Cupertino']}}

Text Classification

# Single-label classification
result = extractor.classify_text(
    "This laptop has amazing performance but terrible battery life!",
    {"sentiment": ["positive", "negative", "neutral"]}
)
print(result)
# Output: {'sentiment': 'negative'}

# Multi-label classification
result = extractor.classify_text(
    "Great camera quality, decent performance, but poor battery life.",
    {
        "aspects": {
            "labels": ["camera", "performance", "battery", "display", "price"],
            "multi_label": True,
            "cls_threshold": 0.4
        }
    }
)
print(result)
# Output: {'aspects': ['camera', 'performance', 'battery']}

Structured Data Extraction

text = "iPhone 15 Pro Max with 256GB storage, A17 Pro chip, priced at $1199."

result = extractor.extract_json(
    text,
    {
        "product": [
            "name::str::Full product name and model",
            "storage::str::Storage capacity",
            "processor::str::Chip or processor information",
            "price::str::Product price with currency"
        ]
    }
)

print(result)
# Output: {
#     'product': [{
#         'name': 'iPhone 15 Pro Max',
#         'storage': '256GB',
#         'processor': 'A17 Pro chip',
#         'price': '$1199'
#     }]
# }

Multi-Task Schema Composition

# Combine all extraction types
schema = (extractor.create_schema()
    .entities({
        "person": "Names of people or individuals",
        "company": "Organization or business names",
        "product": "Products or services mentioned"
    })
    .classification("sentiment", ["positive", "negative", "neutral"])
    .structure("product_info")
        .field("name", dtype="str")
        .field("price", dtype="str")
        .field("features", dtype="list")
)

text = "Apple CEO Tim Cook unveiled the iPhone 15 Pro for $999."
results = extractor.extract(text, schema)

print(results)
# Output: {
#     'entities': {'person': ['Tim Cook'], 'company': ['Apple'], 'product': ['iPhone 15 Pro']},
#     'sentiment': 'positive',
#     'product_info': [{'name': 'iPhone 15 Pro', 'price': '$999', 'features': [...]}]
# }

Model Details

• Model Type: Bidirectional Transformer Encoder (BERT-based)
• Parameters: 205M
• Input: Text sequences
• Output: Entities, classifications, and structured data
• Architecture: Based on GLiNER with multi-task extensions
• Training Data: Multi-domain datasets for NER, classification, and structured extraction

Performance

This model is optimized for:

• Fast CPU inference (no GPU required)
• Low latency applications
• Resource-constrained environments
• Multi-task extraction scenarios

Fine-tuning

This model can be fine-tuned locally with the gliner2 trainer (see the training guide). To fine-tune or serve it without managing GPUs, the Fastino API offers hosted training and inference, usually faster than training locally.

Citation

If you use this model in your research, please cite:

@misc{zaratiana2025gliner2efficientmultitaskinformation,
      title={GLiNER2: An Efficient Multi-Task Information Extraction System with Schema-Driven Interface}, 
      author={Urchade Zaratiana and Gil Pasternak and Oliver Boyd and George Hurn-Maloney and Ash Lewis},
      year={2025},
      eprint={2507.18546},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2507.18546}, 
}

License

This project is licensed under the Apache License 2.0.

Links

• Repository: https://github.com/fastino-ai/GLiNER2
• Paper: https://arxiv.org/abs/2507.18546
• Organization: Fastino AI