Let
$\mu $
be a probability measure on
$\mathrm {GL}_d(\mathbb {R})$
, and denote by
$S_n:= g_n \cdots g_1$
the associated random matrix product, where
$g_j$
are i.i.d. with law
$\mu $
. Under the assumptions that
$\mu $
has a finite exponential moment and generates a proximal and strongly irreducible semigroup, we prove a Berry–Esseen bound with the optimal rate
$O(1/\sqrt n)$
for the coefficients of
$S_n$
, settling a long-standing question considered since the fundamental work of Guivarc’h and Raugi. The local limit theorem for the coefficients is also obtained, complementing a recent partial result of Grama, Quint and Xiao.