Introduction
Today, we are excited to announce the release of Qwen3, the latest addition to the Qwen family of large language models. Our flagship model, Qwen3-235B-A22B, achieves competitive results in benchmark evaluations of coding, math, general capabilities, etc., when compared to other top-tier models such as DeepSeek-R1, o1, o3-mini, Grok-3, and Gemini-2.5-Pro. Additionally, the small MoE model, Qwen3-30B-A3B, outcompetes QwQ-32B with 10 times of activated parameters, and even a tiny model like Qwen3-4B can rival the performance of Qwen2.5-72B-Instruct.
Model Specifications
We are open-weighting two MoE models and six dense models:
- MoE Models:
- Qwen3-235B-A22B: 235B total parameters, 22B activated parameters
- Qwen3-30B-A3B: 30B total parameters, 3B activated parameters
- Dense Models:
- Qwen3-32B, Qwen3-14B, Qwen3-8B, Qwen3-4B, Qwen3-1.7B, Qwen3-0.6B
Key Features
Hybrid Thinking Modes
Qwen3 models introduce a hybrid approach to problem-solving. They support two modes:
- Thinking Mode: The model takes time to reason step by step before delivering the final answer, ideal for complex problems.
- Non-Thinking Mode: The model provides quick, near-instant responses, suitable for simpler questions.
Multilingual Support
Qwen3 models support 119 languages and dialects across multiple language families, including:
- Indo-European languages (English, French, German, etc.)
- Sino-Tibetan languages (Chinese, Burmese)
- Afro-Asiatic languages (Arabic, Hebrew)
- And many more language families
Training and Performance
The training process of Qwen3 involves:
- Long chain-of-thought (CoT) cold start
- Reasoning-based reinforcement learning (RL)
- Thinking mode fusion
- General RL
Usage Guidelines
Qwen3 models are available on platforms like Hugging Face, ModelScope, and Kaggle. For deployment, we recommend using frameworks like SGLang and vLLM. For local usage, tools such as Ollama, LMStudio, MLX, llama.cpp, and KTransformers are highly recommended.