Qwen3.6-27B Launches With Dense Architecture and Strong Coding Results
Qwen3.6-27B launches as a dense multimodal open model with 262K context, strong coding results, and performance that beats Qwen's older MoE flagship.
Read article →Expert guides and pricing analysis for Qwen AI.
Qwen3.6-27B launches as a dense multimodal open model with 262K context, strong coding results, and performance that beats Qwen's older MoE flagship.
Read article →Qwen3.6-Max-Preview is Alibaba's new flagship preview model, built for complex tasks with a 256K context window and frontier-level agentic coding.
Read article →Looking for cheaper Qwen tokens? TokenDock offers Qwen API access at up to 40% below retail pricing with USD billing and OpenAI-compatible access.
Read article →Alibaba ATH’s Happy Oyster is a world model for building interactive virtual environments in real time. Here is what it does and why it matters.
Read article →Qwen3.6-35B-A3B is Qwen’s first open-weight 3.6 model, with 35B total parameters, 3B active parameters, and a strong focus on agentic coding.
Read article →Qwen Code has discontinued its free OAuth tier as of April 15, 2026. Here is what changed and what developers need to do next.
Read article →