Let $ f(x) = \log_3 (1+5^x) $ and $ k \geq 0 $. I need to prove that $ \frac{1}{\min \{ f(x), k \}} $ is convex.
My attempt:
Since $ 1+5^x \geq 1$, then $ \log_3 (1+5^x) \geq 0 $. Furthermore, since $ 1+5^x \geq 1$ is monotonically increasing, $ \log_3 (1+5^x) $ all also be so. Thus,
$$ \min \{ \log_3 (1+5^x), k \} \rightarrow \min \{ 1+5^x, 3^k \} $$
So, now I can equivalently try to prove $ \min \{ 1+5^x, 3^k \} $ that looks simpler. I have tried using the convexity definition, but I have not been successful.

If you rewrite $$ \frac{1}{\min\{f(x),k\}}=\max\left\{\frac1{f(x)},\frac1k\right\} $$ then what is left is to prove that $$ \frac{1}{f(x)}=\frac{1}{\log_3(1+5^x)}=\frac{\ln 3}{\ln(1+5^x)}=\frac{\ln 3}{\ln(1+e^{x\ln 5})} $$ is convex (as the max of two convex functions is convex). It can be done by proving first that the function $$ g(t)=\frac{1}{\ln(1+e^t)} $$ is convex (check that $g''(t)\ge 0$), then $$ \frac{1}{f(x)}=\ln 3\cdot g(x\ln 5) $$ is convex too.