The Q-RWKV-6 32B Instruct Preview is the latest variant of the RWKV model developed by Recursal AI, outperforming previous RWKV, State Space, and Liquid AI models in multiple English benchmark tests. This model successfully converted the weights of the Qwen 32B Instruct model to a customized QRWKV6 architecture by replacing existing Transformer attention heads with RWKV-V6 attention heads, a collaboration spearheaded by the Recursal AI team along with the RWKV and EleutherAI open-source community. Key advantages of this model include significant reductions in large-scale computational costs and the environmentally friendly open-source AI technology.