Qwen3.8-27B Becomes the Most Liked Model on Qwen's Hugging Face Page
A community analysis reveals that the newly released Qwen3.8-27B model has rapidly gained popularity, surpassing the long-reigning QwQ-32B to become the most liked model on Qwen's Hugging Face page in less than a month. It is followed closely in preference by the Qwen3.6-35B-A3B model. This trend highlights a strong preference among local LLM enthusiasts for general-purpose models in the 9B and 27B parameter ranges, which offer a sweet spot between performance and local hardware feasibility. It also shows how quickly the community adopts newer iterations over older reasoning-focused models like QwQ-32B. In less than a month, Qwen3.8-27B accumulated more likes than the next four models combined. Meanwhile, the Qwen3.6-35B-A3B model remains highly regarded as a sparse Mixture-of-Experts (MoE) model that uses only 3 billion active parameters during inference.
## BACKGROUND
Hugging Face is a popular platform where AI developers share open-source models, and "likes" serve as a proxy for community adoption and popularity. QwQ-32B is Qwen's reasoning-focused model designed to compete with models like DeepSeek-R1, while the "A3B" suffix in models like Qwen3.6-35B-A3B refers to having only 3 billion active parameters at inference time despite having 35 billion total parameters.