The new @Alibaba_Qwen Qwen3.5-397B-A17B is live on OpenRouter now! This multimodal model uses a hybrid architecture combining linear attention with sparse MoE for higher inference efficiency. Available as both the open weights version and Qwen3.5 Plus with extended 1M context.

The model delivers SOTA-level performance on: - Language understanding and logical reasoning - Code generation with strong agent capabilities - Image and video understanding Compared to Qwen3, 3.5 shows major improvements in both pure-text and multimodal capabilities.