Prior to the DeepSeek upgrade, Venice utilized GLM-4 as the default model for classic chat