Let $$(P_t)$$
(
P
t
)
be the transition semigroup of the Markov family $$(X^x(t))$$
(
X
x
(
t
)
)
defined by SDE $$\begin{aligned} \mathrm{d}X= b(X)\mathrm{d}t + \mathrm{d}Z, \qquad X(0)=x, \end{aligned}$$
d
X
=
b
(
X
)
d
t
+
d
Z
,
X
(
0
)
=
x
,
where $$Z=\left( Z_1, \ldots , Z_d\right) ^*$$
Z
=
Z
1
,
…
,
Z
d
∗
is a system of independent real-valued Lévy processes. Using the Malliavin calculus we establish the following gradient formula $$\begin{aligned} \nabla P_tf(x)= {\mathbb {E}}\, f\left( X^x(t)\right) Y(t,x), \qquad f\in B_b({\mathbb {R}}^d), \end{aligned}$$
∇
P
t
f
(
x
)
=
E
f
X
x
(
t
)
Y
(
t
,
x
)
,
f
∈
B
b
(
R
d
)
,
where the random field Y does not depend on f. Moreover, in the important cylindrical $$\alpha $$
α
-stable case $$\alpha \in (0,2)$$
α
∈
(
0
,
2
)
, where $$Z_1, \ldots , Z_d$$
Z
1
,
…
,
Z
d
are $$\alpha $$
α
-stable processes, we are able to prove sharp $$L^1$$
L
1
-estimates for Y(t, x). Uniform estimates on $$\nabla P_tf(x)$$
∇
P
t
f
(
x
)
are also given.