The Curious Engineer

All blogs / AI / ML

Qwen

qwenlm.github.io/blog

Blog - Qwen

AI / ML

’22’23’24’25’26
Posts
45 from 14 Nov 2022 to 22 Sep 2025
Posting
now and then
Kind
project news
Feed
https://qwenlm.github.io/blog/index.xml

2025 21

  1. Qwen3Guard: Real-time Safety for Your Token Stream
  2. Qwen-Image-Edit: Image Editing with Higher Quality and Efficiency
  3. Qwen-Image: Crafting with Native Text Rendering
  4. GSPO: Towards Scalable Reinforcement Learning for Language Models
  5. Qwen-MT: Where Speed Meets Smart Translation
  6. Qwen3-Coder: Agentic Coding in the World
  7. 简体中文
  8. Time to Speak Some Dialects, Qwen-TTS!
  9. Qwen VLo: From "Understanding" the World to "Depicting" It
  10. Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models
  11. Qwen3: Think Deeper, Act Faster
  12. QVQ-Max: Think with Evidence
  13. Qwen2.5 Omni: See, Hear, Talk, Write, Do It All!
  14. Qwen2.5-VL-32B: Smarter and Lighter
  15. QwQ-32B: Embracing the Power of Reinforcement Learning
  16. ... QwQ-Max-Preview
  17. Qwen2.5-Max: Exploring the Intelligence of Large-scale MoE Model
  18. Qwen2.5-1M: Deploy Your Own Qwen with Context Length up to 1M Tokens
  19. Qwen2.5 VL! Qwen2.5 VL! Qwen2.5 VL!
  20. Global-batch load balance almost free lunch to improve your MoE LLM training
  21. Towards Effective Process Supervision in Mathematical Reasoning

2024 21

  1. QVQ: To See the World with Wisdom
  2. QwQ: Reflect Deeply on the Boundaries of the Unknown
  3. Extending the Context Length to 1M Tokens!
  4. Qwen2.5-Coder Series: Powerful, Diverse, Practical.
  5. Qwen2.5: A Party of Foundation Models!
  6. Qwen2.5-LLM: Extending the boundary of LLMs
  7. Qwen2.5-Coder: Code More, Learn More!
  8. Qwen2.5-Math: The world's leading open-sourced mathematical LLMs
  9. Qwen2-VL: To See the World More Clearly
  10. Qwen2-Audio: Chat with Your Voice!
  11. Introducing Qwen2-Math
  12. Hello Qwen2
  13. Generalizing an LLM from 8k to 1M Context using Qwen-Agent
  14. Notes on Qwen-Max-0428
  15. Qwen1.5-110B: The First 100B+ Model of the Qwen1.5 Series
  16. Code with CodeQwen1.5
  17. Qwen1.5-32B: Fitting the Capstone of the Qwen1.5 Language Model Series
  18. Qwen1.5-MoE: Matching 7B Model Performance with 1/3 Activated Parameters
  19. Introducing Qwen1.5
  20. Introducing Qwen-VL
  21. Introducing Qwen

2022 3

  1. OFASys: Enabling Multitask Learning with One Line of Code!
  2. Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese
  3. OFA: Towards Building a One-For-All Model