# Three-stage pre-training

Qwen3-30B-A3B-Base is the latest generation of large language models in the Qwen series, with many improvements in training data, model architecture, and optimization techniques, providing more powerful language processing capabilities.

Qwen3-14B-Base is the latest generation of the Tongyi series of large language models, providing a comprehensive set of dense and mixture-of-experts (MoE) models with significant improvements in training data, model architecture, and optimization techniques.

Qwen3-1.7B-Base is the latest generation of large language models in the Tongyi series, offering a range of dense models and mixture-of-experts (MoE) models. It has made significant improvements in training data, model architecture, and optimization techniques.

Qwen3-0.6B-Base is the latest generation of large language models in the Tongyi series. It has a parameter scale of 0.6B, supports 119 languages, and has a context length of up to 32,768 tokens.

Qwen3-0.6B-Base is the latest generation of large language models in the Tongyi Qianwen series, offering a range of dense models and Mixture of Experts (MoE) models.

Qwen3-1.7B is the latest 1.7 billion parameter base language model in the Qwen series, featuring a three-stage pre-training system and supporting 32k context length.