The significant variations in shape and size， blurred boundaries， and complex tissue structures of glioma in Magnetic Resonance Imaging （MRI） Images， lead to challenges in task for brain tumor segmentation. Typically， this task requires researchers with deep professional knowledge to design complex personalized network models， which is time-consuming and needs many human resources. To simplify the network design process and obtain the optimal network architecture automatically， a Neural Architecture Search for multi-tissue segmentation using Convolutional and Transformer-based Networks in glioma segmentation （NASCT-Net） was proposed to improve segmentation precision during construction of network architecture for multi-modal MRI brain tumor segmentation. Firstly， Neural Architecture Search （NAS） technology was applied to construction of an encoder， thereby forming stackable Neural Architecture Search ENcoder CNN （NAS-CNN） encoding modules to optimize the network structure for precise glioma segmentation automatically. Secondly， a feature encoding module based on Transformer was integrated at the lower layers of the encoder to improve the representation ability of relative positions and global information among tumor components. Finally， a Volume-Weighted Dice Loss function （VWDiceLoss） was constructed to address the imbalance between foreground and background. NASCT-Net was compared with methods such as Swin-Unet on BraTS2019 brain tumor datasets. Experimental results show that NASCT-Net has an average Dice Similarity Coefficient （DSC） improvement of 0.009 and an average Hausdorff Distance （HD） decrease of 1.831 mm， validating the effectiveness of NASCT-Net in enhancing the precision of multi-tissue brain tumor segmentation.