#1919th of 90 in Open models

gliner

Small zero-shot named entity recognition models with training and serving code

Stars
4.3k
License
Apache-2.0
Last commit
Nov 2026
Last release
Nov 2026
Language
Python
Stars, last 29 weeks+1,188
25 Apr 2026: 3,084 stars2 May 2026: 3,113 stars9 May 2026: 3,141 stars16 May 2026: 3,169 stars23 May 2026: 3,191 stars30 May 2026: 3,213 stars6 Jun 2026: 3,236 stars13 Jun 2026: 3,264 stars20 Jun 2026: 3,296 stars27 Jun 2026: 3,323 stars4 Jul 2026: 3,348 stars11 Jul 2026: 3,373 stars18 Jul 2026: 3,394 stars25 Jul 2026: 3,423 stars1 Aug 2026: 3,485 stars8 Aug 2026: 3,513 stars15 Aug 2026: 3,534 stars22 Aug 2026: 3,561 stars29 Aug 2026: 3,586 stars5 Sept 2026: 3,609 stars12 Sept 2026: 3,639 stars19 Sept 2026: 3,844 stars26 Sept 2026: 3,956 stars3 Oct 2026: 4,030 stars10 Oct 2026: 4,076 stars17 Oct 2026: 4,128 stars24 Oct 2026: 4,162 stars31 Oct 2026: 4,194 stars7 Nov 2026: 4,237 stars14 Nov 2026: 4,272 stars

3,084 on 25 Apr 20264,272 on 14 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

GLiNER is a Python framework for training and running small transformer models that extract entities of any user-supplied label from text, without task-specific training. Variants cover bi-encoder label scaling, joint entity and relation extraction, streaming NER and PII detection. It runs on CPU, supports FP16/INT8 quantization, torch.compile and ONNX export, and includes a Ray Serve HTTP server.

Who it is for: Engineers needing local entity extraction or PII redaction without an LLM

Strengths

  • Arbitrary entity labels at inference time, no labeled data needed
  • Runs on CPU; FP16, INT8, torch.compile and ONNX export available
  • Several architectures: uni-encoder, bi-encoder, relation extraction, decoder, streaming
  • Includes fine-tuning script and Ray Serve server with dynamic batching

Weaknesses

  • Uni-encoder handles only about 50 entity types; more needs the bi-encoder
  • INT8 accuracy requires models trained with quantization-aware training
  • Classification, entity linking and schema extraction live in separate projects
  • Serving layer depends on Ray Serve, an extra install

What it needs

  • GPU optional
  • Compose
  • Needs PyTorch, Ray Serve (optional, for serving)
  • Models: GLiNER uni-encoder, GLiNER bi-encoder, GLiNER RelEx, GLiNER Decoder, StreamingSpan
  • port 8000

What changed

  1. “FP8 inference requires torchao and a CUDA GPU with compute capability 8.9 or newer. Use compilation for the intended optimized execution path.”

  2. “Install gliner[onnx] for CPU ONNX Runtime, gliner[gpu] for GPU ONNX Runtime, or gliner[openvino] for OpenVINO. Install the separate onnx package when exporting to ONNX.”

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Open models

See all 90
Also in Open models
RankProjectScore
1ruviewWiFi CSI sensing on ESP32 for presence, breathing, heart rate and pose104.5k starsMIT94 out of 100
2layaSingle-pass multilingual classifier returning typed decisions with calibrated odds Live demo ↗ (opens in a new tab)35.6k starsApache-2.088 out of 100
3rf-detrReal-time transformer models for object detection, segmentation and keypoints Live demo ↗ (opens in a new tab)10.1k starsApache-2.082 out of 100
#4timesfmPretrained time-series foundation model for zero-shot forecasting from Google Research35.6k starsApache-2.080 out of 100
#5open_clipPyTorch library for training and running CLIP-style image-text models14.3k starscustom license79 out of 100