The collaboration between startups and large enterprises is a topic of growing interest in the scientific literature. However, although existing research has produced several contributions on this topic, the knowledge in the field appears fragmented. This fragmentation can be traced back to the lack of reference frameworks that clarify the most used concepts, thus providing a shared language. This paper aims to systematically explore the dynamic interplay of asymmetric collaborations between startups and large corporations involved in open innovation projects. The systematic approach adopted in this study leverages a text mining technique, namely Latent Dirichlet Allocation, to synthesize existing research on this topic, employing the MySLR software platform for comprehensive data analysis. A total of 440 relevant articles were analyzed, providing a broad spectrum of insights into the mechanisms, outcomes, and barriers associated with these collaborations. In particular, we characterized the four relevant topics in the investigated research domain: (1) Strategic Alignment and Market Entry, (2) Innovation Transfer, Corporate Strategy and Academic Partnership, (3) Investment dynamics and entrepreneurs’ role in Venture Success, and (4) Organizational Development and Partnership Management: embracing Open Innovation. Our findings reveal that while these partnerships can drive significant technological and economic advancements, they also pose substantial risks and complications. Startups and large corporations often struggle with the integration and alignment of goals and operations, which can impede collaborative efforts. The results of this study may help managers and policymakers understand the nuanced interactions between startups and large companies, and develop frameworks that support and streamline these integrations. Such policies are crucial for ensuring that collaborations are not only innovative but also sustainable and contribute positively to the broader economic and technological landscape.