In OECD countries, Science, Technology and Innovation (STI) policies were seen as key aspects of coping with the Covid-19 pandemic. Now that the pandemic is over, identifying which policy mix portfolios characterised countries in terms of their non-Covid-19 related and Covid-19 specific STI policies fills a knowledge gap on changes in STI policies induced by exogenous shocks. The descriptive nature of this exercise sheds light on the emergency phase, which was addressed in different ways by countries with similar STI policy portfolios in the last decade before the pandemic. Using information on STI policy initiatives in OECD countries, this paper proposes a multidimensional analysis to classify policy initiatives based on both codes (of innovation policy themes, policy instruments and target beneficiaries) and free text policies’ descriptions. Based on text mining and clustering techniques, the multidimensional analysis highlights semantic similarities between the combinations of codes and terms, making it possible to identify policy mixes that characterise non-Covid-19 related and Covid-19 specific STI policies. The cross-country comparison draws attention to the specific policy mix portfolios implemented by countries during the pandemic. The paper contributes to the literature on innovation policy mix in terms of research methods and results in identifying STI policy portfolios and groups of countries with similar structural composition of their innovation policy portfolios, implementing a range of STI strategies in tackling the pandemic. Policy implications of the findings are discussed, with a forward-looking perspective for the analysis of post-pandemic STI policies.