The Network for Innovative Development was developed as the key output of the IPA cross-border project between Hungary and Serbia, conducted under the title "Cross-border Network of Innovative Development and Knowledge Transfer". The network comprises entities from the sectors of academy (universities, faculties, and scientific institutes), business (profit-making and public organisations and business networks) and government (local, regional and national). In addition to the basic constituents of the Triple Helix model, the network also comprises innovation intermediaries. Innovation intermediaries are organizations or groups within organizations working to enable innovation, either directly by enabling the innovativeness of one or more firms, or indirectly by enhancing the innovative capacity of regions, nations, or sectors. They do so by intermediating on the interorganisational level by creating and nurturing interorganisational networks, and by intermediating on the intercommunity level by conducting and supporting technology development activities in the innovation gap between the business and research communities. The number of innovation intermediaries in the Network is large, and their functions and roles are matching. This article investigates the issue of intermediation and the role of intermediaries in the innovation process. The aim of this article is to review and synthesise the literature in this field, and subsequently develop a typology and framework of the different roles and functions of the intermediation process within. The authors of the articles analyse the networked and potential entities of the intermediary category and identify the classes of intermediaries according to the conformity of their roles and functions. The Serbian model of innovation intermediaries was developed based on the experiences of the developed innovative network, and taking into consideration the specific environment. The model entities, i.e. groups have a uniform identified position and significance in the innovation network, and the toolbox for its definition and monitoring. The primary goal of modelling is achieving higher efficiency and efficacy of the formed innovation network, and overall enhancement of innovation in the region.