thinking machines lab's new model, inkling, has made its debut on openrouter's model board, featuring a robust 41 billion active parameters out of a staggering 975 billion total. designed for general-purpose reasoning, coding, and agentic tool-use systems, inkling aims to enhance performance across a variety of applications.
this introduction signals a significant step forward in multimodal AI capabilities, as inkling's architecture allows it to handle diverse tasks more efficiently. developers and researchers will want to keep a close eye on how inkling compares with existing models in terms of real-world performance and usability.
however, the sheer number of parameters doesn't always guarantee superior outcomes, so some may question whether inkling can truly deliver on its ambitious goals. stay tuned for updates on its performance metrics and user feedback as it rolls out in the coming weeks. @OpenRouter
