The paper presents Comix, a new matrix element generator based on the colour dressed Berends-Giele recursive relations. It introduces two new algorithms for phase space integration, optimized for large multiplicities and colour sampling. The generator uses recursive relations to compute tree-level amplitudes efficiently, reducing computational complexity by decomposing four-gluon vertices into simpler ones. This approach allows for the use of colour-dressed vertices in numerical programs, enabling efficient computation of Standard Model processes. The paper discusses the implementation of these relations in Comix, including multi-threading for improved performance. It also presents new integration techniques for phase space generation, including a recursive algorithm that leverages the structure of Feynman diagrams to decompose phase space into manageable components. The methods are validated by comparing results with existing generators like AMEGIC++ and ALPGEN, demonstrating their effectiveness in handling high-multiplicity processes. The recursive phase space generator (RPG) is introduced as a key component, enabling efficient computation of phase space weights by reusing intermediate results and optimizing the integration process. The paper concludes with a comprehensive comparison of results and highlights the efficiency and scalability of the Comix approach for high-energy physics calculations.The paper presents Comix, a new matrix element generator based on the colour dressed Berends-Giele recursive relations. It introduces two new algorithms for phase space integration, optimized for large multiplicities and colour sampling. The generator uses recursive relations to compute tree-level amplitudes efficiently, reducing computational complexity by decomposing four-gluon vertices into simpler ones. This approach allows for the use of colour-dressed vertices in numerical programs, enabling efficient computation of Standard Model processes. The paper discusses the implementation of these relations in Comix, including multi-threading for improved performance. It also presents new integration techniques for phase space generation, including a recursive algorithm that leverages the structure of Feynman diagrams to decompose phase space into manageable components. The methods are validated by comparing results with existing generators like AMEGIC++ and ALPGEN, demonstrating their effectiveness in handling high-multiplicity processes. The recursive phase space generator (RPG) is introduced as a key component, enabling efficient computation of phase space weights by reusing intermediate results and optimizing the integration process. The paper concludes with a comprehensive comparison of results and highlights the efficiency and scalability of the Comix approach for high-energy physics calculations.