Conversion of plastic waste into hydrogen is a potential solution to address the issues of growing demand for hydrogen and the massive accumulation of plastic waste simultaneously. However, most studies on plastic-to-hydrogen technology selection were based on predetermined plastic waste composition, limiting their applicability to dynamic real-life operations. To address this, this work introduces a flexible optimisation model capable of accommodating varying compositions of plastic waste. With the aid of regression models, the optimisation model can optimise the plastic-to-hydrogen production pathways, considering economic and environmental performances, without the constraints of specific plastic waste types or mixture compositions. Regression models are developed based on the ultimate analysis data (carbon, hydrogen, nitrogen, oxygen, and sulphur content) to estimate hydrogen yield and purity across various pathways. Thereafter, fuzzy optimisation is employed to identify the trade-off between cost and environmental impact. In addition to the selection of optimal plastic waste-to-hydrogen pathways, the model also considers different purification technologies that can improve the hydrogen purity to various extents. The model demonstrated that pyrolysis-steam reforming combined with PSA is capable of achieving hydrogen purity of 99.999 % with a highest overall satisfaction of 0.7141 (equivalent to total cost of 3.43 M$ and emissions of 528,647 kg CO2/y) while pyrolysis-catalytic decomposition is more suitable to produce hydrogen with lower purity (55 %).