Is Del (or Nabla) an operator or a vector? Is Del (or Nabla, $\nabla$) an operator or a vector ?
\begin{equation*}
\nabla\equiv\frac{\partial}{\partial x}\vec{i}+\frac{\partial}{\partial y}\vec{j}+\frac{\partial}{\partial z}\vec{k}
\end{equation*}
In some references of vector analysis and electromagnetism, it is considered as an operator (and noted as $\nabla$), and in other ones, it is considered as a vector (and noted as $\vec\nabla$).
 A: First, let's say that $\nabla$ and $\vec \nabla$ are two equivalent notations for the same "object". This notation is used in the representation of three important vector operators: gradient, curl and divergence.
The gradient operator acts on a scalar differentiable function $f(\vec x)$, where $\vec x \in \mathbb R^n$, and returns a vector:
$$\text{grad} \ f(\vec x) = \nabla f(\vec x) \equiv \sum_{i=1}^n \frac{\partial f (\vec x)}{\partial x_i} \vec e_i $$
where $\{\vec e_i \dots\vec e_n\}$ is an orthogonal basis of $\mathbb R^n$.
The divergence operator acts on a vector field $\vec F(\vec x)$, where $\vec x,\vec F \in \mathbb R^n$, and returns a scalar function:
$$\text{div} \ \vec F(\vec x) = \nabla \cdot \vec F(\vec x) \equiv \sum_{i=1}^n \frac{\partial F_i (\vec x)}{\partial x_i} $$
The curl operator acts on a vector field $\vec F(\vec x)$, where $\vec x,\vec F \in \mathbb R^3$, and returns a vector field:
$$\text{curl} \ \vec F(\vec x) = \nabla \times \vec F(\vec x) = \left(\frac{\partial F_z}{\partial y} - \frac{\partial F_y}{\partial z}\right) \hat i+\left(\frac{\partial F_x}{\partial z} - \frac{\partial F_z}{\partial x}\right) \hat j+ \left(\frac{\partial F_y}{\partial x} - \frac{\partial F_x}{\partial y}\right) \hat k$$
where $\hat i, \hat j, \hat k$ are the unit vectors of the three Cartesian axes. 
Notice that, unlike the gradient and divergence, the curl operator does not generalize simply in $n$ dimensions. Also, the notation $\nabla \times \vec F$ is only a mnemonic device useful when we work in cartesian coordinates: in other coordinate systems, applying $\nabla \times \vec F$ will hold the wrong result.
We should probably also mention the laplacian operator, which is the divergence of the gradient:
$$\nabla^2 f(\vec x) \equiv \text{div} \ (\text{grad} \ f(\vec x)) = \nabla \cdot (\nabla f(\vec x))$$

So, to sum up, $\nabla$ is just a useful notation that is used in the representation of three different vector operators. It turns out that we can often formally manipulate $\nabla$ as if it was a vector, but it is not a vector in the usual sense: $\nabla$ alone is meaningless. 
To see this, just consider one of the fundamental properties of vector spaces: if $v,w$ are elements of the vector space $V$, then $v+w$ is also an element of $V$.
Let's consider the vector space $\mathbb R^n$: what meaning should we give to an expression such as
$$\nabla + \vec x \ ?$$
the answer is: no meaning at all, because $\nabla$ is not a vector. 
A: Both. It's an operator that transforms as a covector under rotations. What this means is that if you rotate the coordinate system the gradient in the new coordinate system, $\nabla'$, can be written as:$$\nabla'_i = \sum_{j} R^{-1}_{ij} \nabla_j,$$ where $R^{-1}$ is the inverse of the rotation matrix, $\nabla$ is the gradient in the original coordinate system, and $\nabla'$ is the gradient in the rotated coordinate system.
A: I hate to play this card, but it depends on the object it acts on (and sometimes who you ask.)  Example: many (professors, collegues, etc.)  will insist on differentiating between writing $\vec{\nabla}$ and $\nabla$ (consider obliging if your grade/ income depends on it.)  In reality, however $\nabla$ is NOT a specific operator, but a convenient mathematical notation.  For instance, one may write $\vec{\nabla}\cdot\vec{j}$ or $\nabla\cdot \vec{j}$ and it "should" be obvious from the notation that the meaning of $\nabla$ in this case is a vector operation whether or not the vector symbol is included over it.   Another example:  one may write $(\vec{v}\cdot\vec\nabla) \vec{j}$ or $\vec{v}\cdot\nabla{\vec {j}}$.  In ether case the same quantity is produced.  I appreciate the latter notation, however, because it highlights the freedom to act the $\nabla$ upon $\vec{j}$ first (producing a matrix) and then act on $\vec{v}$ to get a vector, or to act the $\vec{v}$ on $\nabla$ first (producing a scalar operator) and then act on $\vec{j}$ producing an identical vector.  
A: It's an operator in that it maps vectors from one vector space to vectors in another vector space. The spaces in question here are function spaces. 
If you are familiar with Fourier transformations or power series, you know that arbitrary smooth functions can be written as the sum of polynomials or sines and cosines (there are many more sets of functions we can use, but these will suffice to give you the idea). This means that if I have some smooth function $f(x)$, I can write it as 
$$ f(x) = a_1 sin(\frac{\pi x}{L}) + a_2 sin(\frac{2 \pi x}{L}) + ... + b_1 cos(\frac{\pi x}{L})+ ...  $$
It should not be a huge leap to see then that $f(x)$ can be represented as a vector with entries 
$$ \langle a_1, a_2,..., b_1, ...\rangle. $$
The derivative of $f(x)$ is then another vector in this space since it also has a Fourier transformation. This is the sense in which $\nabla$ is an operator. It takes vectors in a function space to vectors in a function space.
Now, given a function of several variables, we could take derivatives of that fucntion along arbitrary combinations of those variables. For instance, given $f(x,y)$ we can take $\frac{\partial}{\partial y}$ or $\frac{\partial}{\partial x}$, or even do something crazy like define $z=\frac{x}{sin(y)}$ and try take $\frac{d}{d z}$. We won't be doing the last thing...
Instead we focus only on "directional derivatives". First recall that linear combinations of $x$ and $y$ define an arrow in $\mathbb{R}^2$, that is the arrow from the origin to $z=a \hat{x} + b \hat{y}$. 
One question we might want to ask is "what is the space of directional derivatives of $f(x,y)$." It is similar to the space of $z$s, but since we're talking about derivatives we'll write the vectors in $\mathbb{R}^2$ as $ a \frac{\partial}{\partial x} + b \frac{\partial}{\partial y}$. This is the sense in which $\nabla$ is a vector, because the space of directional derivatives of functions of $N$ variables (equally functions on $\mathbb{R}^N$) is $\mathbb{R}^N$. 
So as others have pointed out, it is both an operator and a vector. It is a vector in the space of directional derivatives on functions of $N=3$ variables, and it is an operator on functions of $N=3$ variables.
