GLM-130B: An Open Bilingual Pre-trained Model
FreeOpen-source bilingual large language model with 130B parameters
FreeFree tier
Inputs: textOutputs: text
About GLM-130B: An Open Bilingual Pre-trained Model
GLM-130B is an open-source large language model with 130 billion parameters, pre-trained on a bilingual corpus of English and Chinese. It achieves competitive performance on a range of natural language understanding and generation tasks, and is designed to facilitate research and applications in multilingual settings.
Key Features
130 billion parameters
Bilingual pre-training on English and Chinese
Open source and publicly available
Trained on large-scale diverse data
Competitive performance on multiple NLP benchmarks
Pros & Cons
Pros
- State-of-the-art results on many benchmarks for a model of its size
- Open source allows community access and customization
- Bilingual capability supports both English and Chinese applications
- Strong zero-shot and few-shot performance
Cons
- Large model size (130B parameters) requires significant computational resources for inference
- Limited deployment feasibility on consumer hardware without optimization
- Pre-training data composition may introduce biases common in web-scale corpora
Best For
Language understanding and generationMachine translation between English and ChineseQuestion answering and dialogue systemsText summarization and content creation
FAQ
What is GLM-130B?
GLM-130B is an open-source bilingual (English/Chinese) pre-trained language model with 130 billion parameters, developed by researchers at Tsinghua University and Zhipu AI.
Is GLM-130B free to use?
Yes, the model weights and code are released under an open-source license, making it free to use for research and commercial purposes with attribution.