gliner
Generalist model for NER (Extract any entity types from texts)
Decision gist · record as of 2026-08-14
Yes. GLiNER is actively maintained, has no known vulnerabilities, and offers a practical solution for zero-shot entity extraction without labeled data. The permissive Apache-2.0 license and low install friction make it accessible. Install it if you need flexible NER without task-specific training, or if you want to experiment with entity extraction on diverse entity types. The Ray Serve integration and quantization support make it viable for production, though you should verify performance on your specific hardware and entity types before deploying.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or later; torch and transformers must be installed and compatible with your hardware (CPU or GPU).
- Installation is straightforward with low friction; the package depends on torch, transformers, huggingface_hub, tqdm, onnxruntime, and sentencepiece.
- The project is actively maintained with a recent release and 3537 GitHub stars, indicating solid community adoption and ongoing development.
License · maintenance · safety
Apache-2.0 (permissive) — Apache-2.0 is a permissive license, allowing commercial use, modification, and redistribution with minimal restrictions—suitable for most production and research contexts.
last release 2026-07-24 (21 days) · last repo commit 2026-08-10 · 3,537 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 788,163 downloads/mo, #5,057 on PyPI
Alternatives
Verify before relying
pip install gliner
from gliner import GLiNER
model = GLiNER.from_pretrained("gliner-community/gliner_small-v2.5")
entities = model.predict_entities("John works at Google", ["person", "organization"])- Whether pre-trained models are automatically downloaded on first use or require manual setup.
- Specific performance benchmarks comparing GLiNER to ChatGPT or UniNER on standard NER datasets.
- Memory footprint and inference latency on typical consumer hardware (e.g., CPU-only, 8GB RAM).
- Compatibility and performance with ONNX export on edge devices.
What it is and what it does
GLiNER is a named entity recognition framework designed to extract entity types from text without requiring labeled training data. It uses a bi-encoder architecture that pre-computes label embeddings, enabling zero-shot recognition across many entity types. The framework is optimized for deployment on CPUs and consumer hardware through quantization and torch.compile, and includes support for streaming NER, joint entity-relation extraction, and multilingual PII detection across 100+ languages.
The package provides both inference and fine-tuning capabilities. For production workloads, it offers Ray Serve integration for high-throughput serving with dynamic batching and horizontal scaling. Pre-trained models are available via Hugging Face Hub, and the framework supports multiple architectures (uni-encoder, bi-encoder, RelEx, decoder, streaming) tailored to different use cases. Training on custom data is supported through JSON-based datasets and programmatic APIs.
Use it for
- Extract personal identifiable information (SSN, credit cards, emails) from documents for compliance and redaction workflows.
- Build knowledge graphs by jointly extracting entities and relations from unstructured text in a single pass.
- Tag millions of documents against hundreds of entity types in production pipelines using the bi-encoder architecture.
- Fine-tune on domain-specific corpora (biomedical, legal, financial) with minimal labeled examples.
- Parse queries and documents in 100+ languages to extract structured data for multilingual applications.
- Serve NER models at scale with automatic batching and memory-aware GPU allocation via Ray Serve.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
GLiNER is actively maintained, has no known vulnerabilities, and offers a practical solution for zero-shot entity extraction without labeled data. The permissive Apache-2.0 license and low install friction make it accessible. Install it if you need flexible NER without task-specific training, or if you want to experiment with entity extraction on diverse entity types. The Ray Serve integration and quantization support make it viable for production, though you should verify performance on your specific hardware and entity types before deploying.
Install
gliner on PyPI
Before you install
Installation is straightforward with low friction; the package depends on torch, transformers, huggingface_hub, tqdm, onnxruntime, and sentencepiece. The project is actively maintained with a recent release and 3537 GitHub stars, indicating solid community adoption and ongoing development.
Requires Python 3.10 or later; torch and transformers must be installed and compatible with your hardware (CPU or GPU).
License in practice
Apache-2.0 is a permissive license, allowing commercial use, modification, and redistribution with minimal restrictions—suitable for most production and research contexts.
Quickstart
pip install gliner
from gliner import GLiNER
model = GLiNER.from_pretrained("gliner-community/gliner_small-v2.5")
entities = model.predict_entities("John works at Google", ["person", "organization"])Verify before relying
- Whether pre-trained models are automatically downloaded on first use or require manual setup.
- Specific performance benchmarks comparing GLiNER to ChatGPT or UniNER on standard NER datasets.
- Memory footprint and inference latency on typical consumer hardware (e.g., CPU-only, 8GB RAM).
- Compatibility and performance with ONNX export on edge devices.
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 6 packagestorchtransformershuggingface_hubtqdmonnxruntimesentencepiece |
| Maintenance | Actively maintained 21 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 788,163 / month, #5,057 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: gliner-0.2.28-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “zero-shot named entity recognition”
- glinerGLiNER is a lightweight framework for named entity recognition that…
- gliner2GLiNER2 extracts entities, classifies text, parses structured data,…
- recognizers-text-choiceRecognizes and resolves entities like numbers, units, and date/time…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also gliner2 · flair · ginza · langextract · openmed · argus-redact · probablepeople · stanza · setfit