I cant find matrix multiplication under pyopencl.array what am I missing

[TeaCult]

I want to matrix multiply two vectors like cupy's cp.matmul(a,b)
So far I only got this withotu writing elementwise kernel.
Can you point right direction for me ?
Am I missing an example ?

import pyopencl as cl 
from pyopencl import array as clarr
import pyopencl.clmath
import pyopencl.clrandom
import numpy as np


ctx = cl.create_some_context()
queue = cl.CommandQueue(ctx)
a=clarr.arange(queue,100,dtype=np.int32).reshape(4,25)
b=clarr.arange(queue,100,dtype=np.int32).reshape(25,4)
np.matmul(a,b)

array([[cl.Array(19600, dtype=int32), cl.Array(19900, dtype=int32),
        cl.Array(20200, dtype=int32), cl.Array(20500, dtype=int32)],
       [cl.Array(49600, dtype=int32), cl.Array(50525, dtype=int32),
        cl.Array(51450, dtype=int32), cl.Array(52375, dtype=int32)],
       [cl.Array(79600, dtype=int32), cl.Array(81150, dtype=int32),
        cl.Array(82700, dtype=int32), cl.Array(84250, dtype=int32)],
       [cl.Array(109600, dtype=int32), cl.Array(111775, dtype=int32),
        cl.Array(113950, dtype=int32), cl.Array(116125, dtype=int32)]],
      dtype=object)

a=np.arange(1000).reshape(4,25)
b=np.arange(1000).reshape(25,4)
np.matmul(a,b)

array([[ 19600,  19900,  20200,  20500],
       [ 49600,  50525,  51450,  52375],
       [ 79600,  81150,  82700,  84250],
       [109600, 111775, 113950, 116125]])
       

[inducer]

Out of scope for core pyopencl. Write a custom kernel or use something like https://github.com/hunse/pyopencl_blas.