Cryptocurrencies exhibit high volatility, emphasizing the importance of accurately measuring tail risk in their markets. This research incorporates a threshold-switching mechanism into semi-parametric ES-CAViaR models to capture characteristics such as asymmetry and regime shifts. These enhancements improve the modeling of cryptocurrency tail risks while enabling the joint forecasting of Value-at-Risk (VaR) and Expected Shortfall (ES). The proposed models incorporate two types of functions to address the VaR and ES nexus with the option to use the rolling standard deviation of returns as a short-term volatility proxy as a regressor.  We estimate the parameters and forecast tail risk simultaneously within a Bayesian framework. Using the two largest cryptocurrencies by market capitalization, Bitcoin and Ethereum, we evaluate one-step-ahead forecasting performance over a four-year out-of-sample period with a rolling window approach. Backtests and scoring functions show that the proposed threshold models capture cryptocurrency tail risk more accurately than competing approaches.