diff --git a/keras/src/backend/openvino/excluded_concrete_tests.txt b/keras/src/backend/openvino/excluded_concrete_tests.txt
index 34c1d5ff795..2443f23ee57 100644
--- a/keras/src/backend/openvino/excluded_concrete_tests.txt
+++ b/keras/src/backend/openvino/excluded_concrete_tests.txt
@@ -26,7 +26,7 @@ NumpyDtypeTest::test_inner
 NumpyDtypeTest::test_isfinite
 NumpyDtypeTest::test_isinf
 NumpyDtypeTest::test_isnan
-NumpyDtypeTest::test_linspace
+NumpyDtypeTest::test_log1p
 NumpyDtypeTest::test_logaddexp
 NumpyDtypeTest::test_logspace
 NumpyDtypeTest::test_matmul_
@@ -143,7 +143,6 @@ NumpyTwoInputOpsCorrectnessTest::test_digitize
 NumpyTwoInputOpsCorrectnessTest::test_divide_no_nan
 NumpyTwoInputOpsCorrectnessTest::test_einsum
 NumpyTwoInputOpsCorrectnessTest::test_inner
-NumpyTwoInputOpsCorrectnessTest::test_linspace
 NumpyTwoInputOpsCorrectnessTest::test_logspace
 NumpyTwoInputOpsCorrectnessTest::test_outer
 NumpyTwoInputOpsCorrectnessTest::test_quantile
diff --git a/keras/src/backend/openvino/numpy.py b/keras/src/backend/openvino/numpy.py
index 753c7773e11..8a6c421ee12 100644
--- a/keras/src/backend/openvino/numpy.py
+++ b/keras/src/backend/openvino/numpy.py
@@ -907,12 +907,115 @@ def less_equal(x1, x2):
     return OpenVINOKerasTensor(ov_opset.less_equal(x1, x2).output(0))
 
 
-def linspace(
-    start, stop, num=50, endpoint=True, retstep=False, dtype=None, axis=0
-):
-    raise NotImplementedError(
-        "`linspace` is not supported with openvino backend"
-    )
+def linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=None, axis=0):  
+    if not isinstance(num, int):  
+        raise TypeError("num must be an integer")  
+    if num < 0:  
+        raise ValueError("num must be non-negative")  
+    
+    start_ov = get_ov_output(start)  
+    stop_ov = get_ov_output(stop)  
+    
+    if dtype is None:  
+        ov_dtype = OPENVINO_DTYPES[config.floatx()]  
+        numpy_dtype = np.dtype(config.floatx())  
+    else:  
+        if dtype == "bfloat16":  
+            ov_dtype = OPENVINO_DTYPES["float32"]  
+            numpy_dtype = np.dtype("float32")  
+        else:  
+            ov_dtype = OPENVINO_DTYPES[dtype]  
+            numpy_dtype = np.dtype(dtype)  
+    
+    start_ov = ov_opset.convert(start_ov, ov_dtype).output(0)  
+    stop_ov = ov_opset.convert(stop_ov, ov_dtype).output(0)  
+    
+    start_shape = start_ov.get_shape()  
+    stop_shape = stop_ov.get_shape()  
+    
+    if num == 0:  
+        if len(start_shape) == 0 and len(stop_shape) == 0:  
+            empty_array = np.array([], dtype=numpy_dtype)  
+            result = ov_opset.constant(empty_array).output(0)  
+        else:  
+            broadcast_shape = list(np.broadcast(  
+                np.empty(start_shape, dtype=bool),  
+                np.empty(stop_shape, dtype=bool)  
+            ).shape)  
+            broadcast_shape.insert(axis, 0)  
+            
+            empty_array = np.empty(broadcast_shape, dtype=numpy_dtype)  
+            result = ov_opset.constant(empty_array).output(0)  
+        
+        if retstep:  
+            delta = ov_opset.subtract(stop_ov, start_ov).output(0)  
+            return OpenVINOKerasTensor(result), OpenVINOKerasTensor(delta)  
+        return OpenVINOKerasTensor(result)  
+    
+    if num == 1:  
+        if endpoint:  
+            result = start_ov  
+        else:  
+            result = start_ov  
+        
+        if len(start_shape) > 0 or len(stop_shape) > 0:  
+            broadcast_shape = list(np.broadcast(  
+                np.empty(start_shape, dtype=bool),  
+                np.empty(stop_shape, dtype=bool)  
+            ).shape)  
+            broadcast_shape.insert(axis, 1)  
+            
+            shape_const = ov_opset.constant(np.array(broadcast_shape, dtype=np.int64)).output(0)  
+            result = ov_opset.reshape(result, shape_const).output(0)  
+        
+        if retstep:  
+            delta = ov_opset.subtract(stop_ov, start_ov).output(0)  
+            return OpenVINOKerasTensor(result), OpenVINOKerasTensor(delta)  
+        return OpenVINOKerasTensor(result)  
+    
+    div = num - 1 if endpoint else num
+    div_const = ov_opset.constant(float(div)).output(0)  
+    delta = ov_opset.subtract(stop_ov, start_ov).output(0)  
+    step = ov_opset.divide(delta, div_const).output(0)  
+    
+    indices_np = np.arange(num, dtype=numpy_dtype)  
+    indices = ov_opset.constant(indices_np).output(0)  
+    
+    is_scalar_input = len(start_shape) == 0 and len(stop_shape) == 0  
+    
+    if is_scalar_input:  
+        scaled_indices = ov_opset.multiply(indices, step).output(0)  
+        result = ov_opset.add(start_ov, scaled_indices).output(0)  
+        
+        if axis != 0:  
+            out_shape = [1] * (axis + 1)  
+            out_shape[axis] = num  
+            shape_const = ov_opset.constant(np.array(out_shape, dtype=np.int64)).output(0)  
+            result = ov_opset.reshape(result, shape_const).output(0)  
+    else:  
+        broadcast_shape = list(np.broadcast(  
+            np.empty(start_shape, dtype=bool),  
+            np.empty(stop_shape, dtype=bool)  
+        ).shape)  
+        
+        expanded_shape = broadcast_shape.copy()  
+        expanded_shape.insert(axis, 1)  
+        
+        shape_const = ov_opset.constant(np.array(expanded_shape, dtype=np.int64)).output(0)  
+        start_reshaped = ov_opset.reshape(start_ov, shape_const).output(0)  
+        step_reshaped = ov_opset.reshape(step, shape_const).output(0)  
+        
+        indices_shape = [1] * len(expanded_shape)  
+        indices_shape[axis] = num  
+        indices_shape_const = ov_opset.constant(np.array(indices_shape, dtype=np.int64)).output(0)  
+        indices_reshaped = ov_opset.reshape(indices, indices_shape_const).output(0)  
+        
+        scaled_indices = ov_opset.multiply(indices_reshaped, step_reshaped).output(0)  
+        result = ov_opset.add(start_reshaped, scaled_indices).output(0)  
+    
+    if retstep:  
+        return OpenVINOKerasTensor(result), OpenVINOKerasTensor(step)  
+    return OpenVINOKerasTensor(result)  
 
 
 def log(x):