Assuming the Axiom of Choice, suppose that the arbitrary vector space $V$ has the Hamel basis ${\cal B} = ({\bf e}_i)_{i \in I}$. Because ${\cal B}$ is a basis, the set ${\cal B}$ is a subset of $V$. So, $$\forall i\in I: {\bf e}_i\in V,$$which implies every finite linear combination of ${\bf e}_i$'s belongs to $V$. This means that any eventually-zero sequence $x_n = (a_1,a_2,a_3,\dots,a_n,0,0,0,\dots)$ corresponds to a vector in $V$. Conversely, every vector in $V$ corresponds to a eventually-zero sequence since by the definition of basis, every vector can be written as a finite linear combination of ${\bf e}_i$'s. Now define the norm $\|{\bf x}\| = \max\{|a_i| \mid i \in F \}$ where ${\bf x} =\sum_{i \in F}a_i{\bf e}_i$. It can be proved that $\|{\bf x}\|$ is indeed a well-defined norm. In order to prove that vector space $V$ with this norm, in the case $\dim(V)=\infty$, is an incomplete normed space, consider the sequence, $${\bf x}_n = \sum_{i=1}^n\frac{1}{i}{\bf e}_i \ \ \text{corresponds to} \ \ (\frac{1}{1},\frac{1}{2},\frac{1}{3},\dots,\frac{1}{n},0,0,0,\dots).$$ This sequence is Cauchy since if $m\gt n$ then, $$||{\bf x}_m - {\bf x}_n|| = ||\sum_{i=n+1}^m\frac{1}{i}{\bf e}_i|| = \frac{1}{n+1},$$ and $\frac{1}{n+1}$ tends to $0$ as $n$ and $m$ tend to $+\infty$ but the sequence is not convergent since it has to converge to some ${\bf x} =(a_1,a_2,\dots,a_m,0,0,\dots)$ and for $n \ge m+1$ we have, $$||{\bf x} - {\bf x}_n||\ge \frac{1}{m+1},$$which shows that the sequence is at a positive distance from any element in $V$. Interestingly, it can be shown that the space of eventually-zero sequences is not a complete space under any norm.
Conclusion: In the case $\dim(V) \lt \infty$, the vector space is always complete. If $\dim(V) = \infty$ and $V$ is an incomplete normed vector space, then the basis can be countable or uncountable but if $V$ is a complete normed vector space, then the basis is uncountable.