Qwen3.5 Small: 0.8B, 2B, 4B, 9B Released (huggingface.co)

Qwen has announced the release of the Qwen3.5 model family, which includes both Qwen3.5 Small (featuring configurations of 0.8B, 2B, 4B, and 9B parameters) and Qwen3.5 Medium (offering a much larger scale with models of 35B-A3B, 27B, 122B-A10B, and 397B-A17B). This expansion highlights Qwen's commitment to enhancing image-text-to-text capabilities and improving the versatility of AI models, catering to a wide range of applications in AI/ML. The significance of this release lies in its potential to empower developers and researchers with an array of options that scale according to their specific needs, whether for lightweight applications or resource-intensive tasks. By introducing a spectrum of model sizes, Qwen allows for broader accessibility, enabling performance optimizations and quicker deployment. With these updates, the AI community can expect enhanced performance and efficiency in applications that leverage multimodal data, thereby pushing the boundaries of what's achievable with AI technologies.