diff --git a/tfjs-core/src/gradients/Reverse_grad.ts b/tfjs-core/src/gradients/Reverse_grad.ts
new file mode 100644
index 00000000000..028ebbb8755
--- /dev/null
+++ b/tfjs-core/src/gradients/Reverse_grad.ts
@@ -0,0 +1,30 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+import {Reverse, ReverseAttrs} from '../kernel_names';
+import {GradConfig, NamedAttrMap} from '../kernel_registry';
+import {reverse} from '../ops/reverse';
+import {Tensor} from '../tensor';
+import {parseAxisParam} from '../util';
+
+export const reverseGradConfig: GradConfig = {
+  kernelName: Reverse,
+  gradFunc: (dy: Tensor, saved: Tensor[], attrs: NamedAttrMap) => {
+    const {dims} = attrs as {} as ReverseAttrs;
+    const axes = parseAxisParam(dims, dy.shape);
+    return {x: () => reverse(dy, axes)};
+  }
+};
diff --git a/tfjs-core/src/kernel_names.ts b/tfjs-core/src/kernel_names.ts
index 1a7e4992b94..64c790ddd78 100644
--- a/tfjs-core/src/kernel_names.ts
+++ b/tfjs-core/src/kernel_names.ts
@@ -440,6 +440,12 @@ export type ResizeBilinearGradInputs = Pick<NamedTensorInfoMap, 'images'>;
 export const Relu6 = 'Relu6';
 export type Relu6Inputs = Pick<NamedTensorInfoMap, 'x'>;
 
+export const Reverse = 'Reverse';
+export type ReverseInputs = Pick<NamedTensorInfoMap, 'x'>;
+export interface ReverseAttrs {
+  dims: number|number[];
+}
+
 export const SelectV2 = 'SelectV2';
 export type SelectV2Inputs = Pick<NamedTensorInfoMap, 'condition'|'t'|'e'>;
 
diff --git a/tfjs-core/src/ops/ops.ts b/tfjs-core/src/ops/ops.ts
index f4a236ade57..5ba687ee716 100644
--- a/tfjs-core/src/ops/ops.ts
+++ b/tfjs-core/src/ops/ops.ts
@@ -93,6 +93,11 @@ export {randomUniform} from './random_uniform';
 export {real} from './real';
 export {relu} from './relu';
 export {relu6} from './relu6';
+export {reverse} from './reverse';
+export {reverse1d} from './reverse_1d';
+export {reverse2d} from './reverse_2d';
+export {reverse3d} from './reverse_3d';
+export {reverse4d} from './reverse_4d';
 export {selu} from './selu';
 export {separableConv2d} from './separable_conv2d';
 export {spaceToBatchND} from './space_to_batch_nd';
@@ -106,7 +111,6 @@ export {where} from './where';
 export {whereAsync} from './where_async';
 
 export * from './boolean_mask';
-export * from './reverse';
 export * from './slice';
 export * from './unary_ops';
 export * from './reduction_ops';
diff --git a/tfjs-core/src/ops/reverse.ts b/tfjs-core/src/ops/reverse.ts
index 1e72f4af8da..875d8978e88 100644
--- a/tfjs-core/src/ops/reverse.ts
+++ b/tfjs-core/src/ops/reverse.ts
@@ -15,70 +15,18 @@
  * =============================================================================
  */
 
-import {ENGINE} from '../engine';
-import {Tensor, Tensor1D, Tensor2D, Tensor3D, Tensor4D} from '../tensor';
+import {ENGINE, ForwardFunc} from '../engine';
+import {Reverse, ReverseAttrs, ReverseInputs} from '../kernel_names';
+import {NamedAttrMap} from '../kernel_registry';
+import {Tensor} from '../tensor';
+import {NamedTensorMap} from '../tensor_types';
 import {convertToTensor} from '../tensor_util_env';
 import {TensorLike} from '../types';
-import * as util from '../util';
-import {op} from './operation';
+import {parseAxisParam} from '../util';
 
-/**
- * Reverses a `tf.Tensor1D`.
- *
- * @param x The input tensor.
- */
-function reverse1d_(x: Tensor1D|TensorLike): Tensor1D {
-  const $x = convertToTensor(x, 'x', 'reverse');
-  util.assert(
-      $x.rank === 1,
-      () => `Error in reverse1D: x must be rank 1 but got rank ${$x.rank}.`);
-  return reverse($x, 0);
-}
-
-/**
- * Reverses a `tf.Tensor2D` along a specified axis.
- *
- * @param x The input tensor.
- * @param axis The set of dimensions to reverse. Must be in the
- *     range [-rank(x), rank(x)). Defaults to all axes.
- */
-function reverse2d_(x: Tensor2D|TensorLike, axis?: number|number[]): Tensor2D {
-  const $x = convertToTensor(x, 'x', 'reverse');
-  util.assert(
-      $x.rank === 2,
-      () => `Error in reverse2D: x must be rank 2 but got rank ${$x.rank}.`);
-  return reverse($x, axis);
-}
-
-/**
- * Reverses a `tf.Tensor3D` along a specified axis.
- *
- * @param x The input tensor.
- * @param axis The set of dimensions to reverse. Must be in the
- *     range [-rank(x), rank(x)). Defaults to all axes.
- */
-function reverse3d_(x: Tensor3D|TensorLike, axis?: number|number[]): Tensor3D {
-  const $x = convertToTensor(x, 'x', 'reverse');
-  util.assert(
-      $x.rank === 3,
-      () => `Error in reverse3D: x must be rank 3 but got rank ${$x.rank}.`);
-  return reverse($x, axis);
-}
-
-/**
- * Reverses a `tf.Tensor4D` along a specified axis.
- *
- * @param x The input tensor.
- * @param axis The set of dimensions to reverse. Must be in the
- *     range [-rank(x), rank(x)). Defaults to all axes.
- */
-function reverse4d_(x: Tensor4D|TensorLike, axis?: number|number[]): Tensor4D {
-  const $x = convertToTensor(x, 'x', 'reverse');
-  util.assert(
-      $x.rank === 4,
-      () => `Error in reverse4D: x must be rank 4 but got rank ${$x.rank}.`);
-  return reverse($x, axis);
-}
+import {reshape} from './array_ops';
+import {clone} from './clone';
+import {op} from './operation';
 
 /**
  * Reverses a `tf.Tensor` along a specified axis.
@@ -114,20 +62,21 @@ function reverse_<T extends Tensor>(
     x: T|TensorLike, axis?: number|number[]): T {
   const $x = convertToTensor(x, 'x', 'reverse');
 
-  if ($x.rank === 0) {
-    return $x.clone();
-  }
-  const axes = util.parseAxisParam(axis, $x.shape);
-  const grad = (dy: T) => {
-    return {$x: () => dy.reverse(axes)};
+  const forward: ForwardFunc<Tensor> = (backend) => {
+    const axes = parseAxisParam(axis, $x.shape);
+    if ($x.rank === 0) {
+      return clone($x);
+    }
+    const res = backend.reverse($x, axes);
+    return reshape(res, $x.shape);
   };
-  const res =
-      ENGINE.runKernelFunc(backend => backend.reverse($x, axes), {$x}, grad);
-  return res.reshapeAs($x);
+
+  const inputs: ReverseInputs = {x: $x};
+  const attrs: ReverseAttrs = {dims: axis};
+
+  return ENGINE.runKernelFunc(
+             forward, inputs as {} as NamedTensorMap, null /* gradient */,
+             Reverse, attrs as {} as NamedAttrMap) as T;
 }
 
 export const reverse = op({reverse_});
-export const reverse1d = op({reverse1d_});
-export const reverse2d = op({reverse2d_});
-export const reverse3d = op({reverse3d_});
-export const reverse4d = op({reverse4d_});
diff --git a/tfjs-core/src/ops/reverse_1d.ts b/tfjs-core/src/ops/reverse_1d.ts
new file mode 100644
index 00000000000..2c980242e0c
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_1d.ts
@@ -0,0 +1,38 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import {Tensor1D} from '../tensor';
+import {convertToTensor} from '../tensor_util_env';
+import {TensorLike} from '../types';
+import * as util from '../util';
+import {op} from './operation';
+import {reverse} from './reverse';
+
+/**
+ * Reverses a `tf.Tensor1D`.
+ *
+ * @param x The input tensor.
+ */
+function reverse1d_(x: Tensor1D|TensorLike): Tensor1D {
+  const $x = convertToTensor(x, 'x', 'reverse');
+  util.assert(
+      $x.rank === 1,
+      () => `Error in reverse1D: x must be rank 1 but got rank ${$x.rank}.`);
+  return reverse($x, 0);
+}
+
+export const reverse1d = op({reverse1d_});
diff --git a/tfjs-core/src/ops/reverse_1d_test.ts b/tfjs-core/src/ops/reverse_1d_test.ts
new file mode 100644
index 00000000000..4e11f70be7c
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_1d_test.ts
@@ -0,0 +1,60 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import * as tf from '../index';
+import {ALL_ENVS, describeWithFlags} from '../jasmine_util';
+import {expectArraysClose} from '../test_util';
+
+describeWithFlags('reverse1d', ALL_ENVS, () => {
+  it('reverse a 1D array', async () => {
+    const input = tf.tensor1d([1, 2, 3, 4, 5]);
+    const result = tf.reverse1d(input);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [5, 4, 3, 2, 1]);
+  });
+
+  it('reverse a 1D array, even length', async () => {
+    const input = tf.tensor1d([1, 2, 3, 4]);
+    const result = tf.reverse1d(input);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [4, 3, 2, 1]);
+  });
+
+  it('grad', async () => {
+    const a = tf.tensor1d([1, 2, 3]);
+    const dy = tf.tensor1d([10, 20, 30]);
+    const da = tf.grad((a: tf.Tensor1D) => tf.reverse1d(a))(a, dy);
+    expect(da.shape).toEqual([3]);
+    expectArraysClose(await da.data(), [30, 20, 10]);
+  });
+
+  it('gradient with clones', async () => {
+    const a = tf.tensor1d([1, 2, 3]);
+    const dy = tf.tensor1d([10, 20, 30]);
+    const da =
+        tf.grad((a: tf.Tensor1D) => tf.reverse1d(a.clone()).clone())(a, dy);
+    expect(da.shape).toEqual([3]);
+    expectArraysClose(await da.data(), [30, 20, 10]);
+  });
+
+  it('accepts a tensor-like object', async () => {
+    const input = [1, 2, 3, 4, 5];
+    const result = tf.reverse1d(input);
+    expect(result.shape).toEqual([5]);
+    expectArraysClose(await result.data(), [5, 4, 3, 2, 1]);
+  });
+});
diff --git a/tfjs-core/src/ops/reverse_2d.ts b/tfjs-core/src/ops/reverse_2d.ts
new file mode 100644
index 00000000000..96fe4d7e05f
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_2d.ts
@@ -0,0 +1,40 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import {Tensor2D} from '../tensor';
+import {convertToTensor} from '../tensor_util_env';
+import {TensorLike} from '../types';
+import * as util from '../util';
+import {op} from './operation';
+import {reverse} from './reverse';
+
+/**
+ * Reverses a `tf.Tensor2D` along a specified axis.
+ *
+ * @param x The input tensor.
+ * @param axis The set of dimensions to reverse. Must be in the
+ *     range [-rank(x), rank(x)). Defaults to all axes.
+ */
+function reverse2d_(x: Tensor2D|TensorLike, axis?: number|number[]): Tensor2D {
+  const $x = convertToTensor(x, 'x', 'reverse');
+  util.assert(
+      $x.rank === 2,
+      () => `Error in reverse2D: x must be rank 2 but got rank ${$x.rank}.`);
+  return reverse($x, axis);
+}
+
+export const reverse2d = op({reverse2d_});
diff --git a/tfjs-core/src/ops/reverse_2d_test.ts b/tfjs-core/src/ops/reverse_2d_test.ts
new file mode 100644
index 00000000000..30b69a2c164
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_2d_test.ts
@@ -0,0 +1,102 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import * as tf from '../index';
+import {ALL_ENVS, describeWithFlags} from '../jasmine_util';
+import {expectArraysClose} from '../test_util';
+
+describeWithFlags('reverse2d', ALL_ENVS, () => {
+  it('reverse a 2D array at axis [0]', async () => {
+    const axis = [0];
+    const a = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
+    const result = tf.reverse2d(a, axis);
+
+    expect(result.shape).toEqual(a.shape);
+    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
+  });
+
+  it('reverse a 2D array at axis [1]', async () => {
+    const axis = [1];
+    const a = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
+    const result = tf.reverse2d(a, axis);
+
+    expect(result.shape).toEqual(a.shape);
+    expectArraysClose(await result.data(), [3, 2, 1, 6, 5, 4]);
+  });
+
+  it('reverse a 2D array odd rows and columns at axis [0, 1]', async () => {
+    const axis = [0, 1];
+    const a = tf.tensor2d(
+        [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], [3, 5]);
+    const result = tf.reverse2d(a, axis);
+
+    expect(result.shape).toEqual(a.shape);
+    expectArraysClose(
+        await result.data(),
+        [15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1]);
+  });
+
+  it('throws error with invalid input', () => {
+    // tslint:disable-next-line:no-any
+    const x: any = tf.tensor1d([1, 20, 300, 4]);
+    expect(() => tf.reverse2d(x, [0])).toThrowError();
+  });
+
+  it('throws error with invalid axis param', () => {
+    const x = tf.tensor2d([1, 20, 300, 4], [1, 4]);
+    expect(() => tf.reverse2d(x, [2])).toThrowError();
+    expect(() => tf.reverse2d(x, [-3])).toThrowError();
+  });
+
+  it('throws error with non integer axis param', () => {
+    const x = tf.tensor2d([1, 20, 300, 4], [1, 4]);
+    expect(() => tf.reverse2d(x, [0.5])).toThrowError();
+  });
+
+  it('grad', async () => {
+    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
+    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
+    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a))(a, dy);
+    expect(da.shape).toEqual([2, 3]);
+    expectArraysClose(await da.data(), [60, 50, 40, 30, 20, 10]);
+  });
+
+  it('grad with reverse(axis=0)', async () => {
+    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
+    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
+    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a, 0))(a, dy);
+    expect(da.shape).toEqual([2, 3]);
+    expectArraysClose(await da.data(), [40, 50, 60, 10, 20, 30]);
+  });
+
+  it('grad with reverse(axis=1)', async () => {
+    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
+    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
+    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a, 1))(a, dy);
+    expect(da.shape).toEqual([2, 3]);
+    expectArraysClose(await da.data(), [30, 20, 10, 60, 50, 40]);
+  });
+
+  it('accepts a tensor-like object', async () => {
+    const axis = [0];
+    const a = [[1, 2, 3], [4, 5, 6]];  // 2x3
+    const result = tf.reverse2d(a, axis);
+
+    expect(result.shape).toEqual([2, 3]);
+    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
+  });
+});
diff --git a/tfjs-core/src/ops/reverse_3d.ts b/tfjs-core/src/ops/reverse_3d.ts
new file mode 100644
index 00000000000..9b6e4615513
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_3d.ts
@@ -0,0 +1,40 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import {Tensor3D} from '../tensor';
+import {convertToTensor} from '../tensor_util_env';
+import {TensorLike} from '../types';
+import * as util from '../util';
+import {op} from './operation';
+import {reverse} from './reverse';
+
+/**
+ * Reverses a `tf.Tensor3D` along a specified axis.
+ *
+ * @param x The input tensor.
+ * @param axis The set of dimensions to reverse. Must be in the
+ *     range [-rank(x), rank(x)). Defaults to all axes.
+ */
+function reverse3d_(x: Tensor3D|TensorLike, axis?: number|number[]): Tensor3D {
+  const $x = convertToTensor(x, 'x', 'reverse');
+  util.assert(
+      $x.rank === 3,
+      () => `Error in reverse3D: x must be rank 3 but got rank ${$x.rank}.`);
+  return reverse($x, axis);
+}
+
+export const reverse3d = op({reverse3d_});
diff --git a/tfjs-core/src/ops/reverse_3d_test.ts b/tfjs-core/src/ops/reverse_3d_test.ts
new file mode 100644
index 00000000000..0b6216347a4
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_3d_test.ts
@@ -0,0 +1,124 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import * as tf from '../index';
+import {ALL_ENVS, describeWithFlags} from '../jasmine_util';
+import {expectArraysClose} from '../test_util';
+
+describeWithFlags('reverse3d', ALL_ENVS, () => {
+  // [
+  //   [
+  //     [0,  1,  2,  3],
+  //     [4,  5,  6,  7],
+  //     [8,  9,  10, 11]
+  //   ],
+  //   [
+  //     [12, 13, 14, 15],
+  //     [16, 17, 18, 19],
+  //     [20, 21, 22, 23]
+  //   ]
+  // ]
+  const shape: [number, number, number] = [2, 3, 4];
+  const data = [
+    0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11,
+    12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
+  ];
+
+  it('reverse a 3D array at axis [0]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [0]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
+      0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11
+    ]);
+  });
+
+  it('reverse a 3D array at axis [1]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [1]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3,
+      20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15
+    ]);
+  });
+
+  it('reverse a 3D array at axis [2]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [2]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8,
+      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20
+    ]);
+  });
+
+  it('reverse a 3D array at axis [0, 1]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [0, 1]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15,
+      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3
+    ]);
+  });
+
+  it('reverse a 3D array at axis [0, 2]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [0, 2]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20,
+      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8
+    ]);
+  });
+
+  it('reverse a 3D array at axis [1, 2]', async () => {
+    const input = tf.tensor3d(data, shape);
+    const result = tf.reverse3d(input, [1, 2]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      11, 10, 9,  8,  7,  6,  5,  4,  3,  2,  1,  0,
+      23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12
+    ]);
+  });
+
+  it('throws error with invalid input', () => {
+    // tslint:disable-next-line:no-any
+    const x: any = tf.tensor2d([1, 20, 300, 4], [1, 4]);
+    expect(() => tf.reverse3d(x, [1])).toThrowError();
+  });
+
+  it('throws error with invalid axis param', () => {
+    const x = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
+    expect(() => tf.reverse3d(x, [3])).toThrowError();
+    expect(() => tf.reverse3d(x, [-4])).toThrowError();
+  });
+
+  it('throws error with non integer axis param', () => {
+    const x = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
+    expect(() => tf.reverse3d(x, [0.5])).toThrowError();
+  });
+
+  it('accepts a tensor-like object', async () => {
+    const input = [[[1], [2], [3]], [[4], [5], [6]]];  // 2x3x1
+    const result = tf.reverse3d(input, [0]);
+    expect(result.shape).toEqual([2, 3, 1]);
+    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
+  });
+});
diff --git a/tfjs-core/src/ops/reverse_4d.ts b/tfjs-core/src/ops/reverse_4d.ts
new file mode 100644
index 00000000000..72554eb5327
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_4d.ts
@@ -0,0 +1,40 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import {Tensor4D} from '../tensor';
+import {convertToTensor} from '../tensor_util_env';
+import {TensorLike} from '../types';
+import * as util from '../util';
+import {op} from './operation';
+import {reverse} from './reverse';
+
+/**
+ * Reverses a `tf.Tensor4D` along a specified axis.
+ *
+ * @param x The input tensor.
+ * @param axis The set of dimensions to reverse. Must be in the
+ *     range [-rank(x), rank(x)). Defaults to all axes.
+ */
+function reverse4d_(x: Tensor4D|TensorLike, axis?: number|number[]): Tensor4D {
+  const $x = convertToTensor(x, 'x', 'reverse');
+  util.assert(
+      $x.rank === 4,
+      () => `Error in reverse4D: x must be rank 4 but got rank ${$x.rank}.`);
+  return reverse($x, axis);
+}
+
+export const reverse4d = op({reverse4d_});
diff --git a/tfjs-core/src/ops/reverse_4d_test.ts b/tfjs-core/src/ops/reverse_4d_test.ts
new file mode 100644
index 00000000000..7e46cc4767e
--- /dev/null
+++ b/tfjs-core/src/ops/reverse_4d_test.ts
@@ -0,0 +1,164 @@
+/**
+ * @license
+ * Copyright 2020 Google Inc. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+
+import * as tf from '../index';
+import {ALL_ENVS, describeWithFlags} from '../jasmine_util';
+import {expectArraysClose} from '../test_util';
+
+describeWithFlags('reverse4d', ALL_ENVS, () => {
+  // [
+  //   [
+  //     [
+  //       [0,  1,  2,  3],
+  //       [4,  5,  6,  7],
+  //       [8,  9,  10, 11]
+  //     ],
+  //     [
+  //       [12, 13, 14, 15],
+  //       [16, 17, 18, 19],
+  //       [20, 21, 22, 23]
+  //     ]
+  //   ],
+  //   [
+  //     [
+  //       [24, 25, 26, 27],
+  //       [28, 29, 30, 31],
+  //       [32, 33, 34, 35]
+  //     ],
+  //     [
+  //       [36, 37, 38, 39],
+  //       [40, 41, 42, 43],
+  //       [44, 45, 46, 47]
+  //     ]
+  //   ],
+  //   [
+  //     [
+  //       [48, 49, 50, 51],
+  //       [52, 53, 54, 55],
+  //       [56, 57, 58, 59]
+  //     ],
+  //     [
+  //       [60, 61, 62, 63],
+  //       [64, 65, 66, 67],
+  //       [68, 69, 70, 71]
+  //     ]
+  //   ]
+  // ]
+  const shape: [number, number, number, number] = [3, 2, 3, 4];
+  const data = [
+    0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11, 12, 13, 14, 15, 16, 17,
+    18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
+    36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
+    54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71
+  ];
+
+  it('reverse a 4D array at axis [0]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [0]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65,
+      66, 67, 68, 69, 70, 71, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
+      36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 0,  1,  2,  3,  4,  5,
+      6,  7,  8,  9,  10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
+    ]);
+  });
+
+  it('reverse a 4D array at axis [1]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [1]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 0,  1,  2,  3,  4,  5,
+      6,  7,  8,  9,  10, 11, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47,
+      24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 60, 61, 62, 63, 64, 65,
+      66, 67, 68, 69, 70, 71, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59
+    ]);
+  });
+
+  it('reverse a 4D array at axis [2]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [2]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3,  20, 21, 22, 23, 16, 17,
+      18, 19, 12, 13, 14, 15, 32, 33, 34, 35, 28, 29, 30, 31, 24, 25, 26, 27,
+      44, 45, 46, 47, 40, 41, 42, 43, 36, 37, 38, 39, 56, 57, 58, 59, 52, 53,
+      54, 55, 48, 49, 50, 51, 68, 69, 70, 71, 64, 65, 66, 67, 60, 61, 62, 63
+    ]);
+  });
+
+  it('reverse a 4D array at axis [3]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [3]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8,  15, 14, 13, 12, 19, 18,
+      17, 16, 23, 22, 21, 20, 27, 26, 25, 24, 31, 30, 29, 28, 35, 34, 33, 32,
+      39, 38, 37, 36, 43, 42, 41, 40, 47, 46, 45, 44, 51, 50, 49, 48, 55, 54,
+      53, 52, 59, 58, 57, 56, 63, 62, 61, 60, 67, 66, 65, 64, 71, 70, 69, 68
+    ]);
+  });
+
+  it('reverse a 4D array at axis [0, 2]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [0, 2]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      56, 57, 58, 59, 52, 53, 54, 55, 48, 49, 50, 51, 68, 69, 70, 71, 64, 65,
+      66, 67, 60, 61, 62, 63, 32, 33, 34, 35, 28, 29, 30, 31, 24, 25, 26, 27,
+      44, 45, 46, 47, 40, 41, 42, 43, 36, 37, 38, 39, 8,  9,  10, 11, 4,  5,
+      6,  7,  0,  1,  2,  3,  20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15
+    ]);
+  });
+
+  it('reverse a 4D array at axis [1, 3]', async () => {
+    const input = tf.tensor4d(data, shape);
+    const result = tf.reverse4d(input, [1, 3]);
+    expect(result.shape).toEqual(input.shape);
+    expectArraysClose(await result.data(), [
+      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20, 3,  2,  1,  0,  7,  6,
+      5,  4,  11, 10, 9,  8,  39, 38, 37, 36, 43, 42, 41, 40, 47, 46, 45, 44,
+      27, 26, 25, 24, 31, 30, 29, 28, 35, 34, 33, 32, 63, 62, 61, 60, 67, 66,
+      65, 64, 71, 70, 69, 68, 51, 50, 49, 48, 55, 54, 53, 52, 59, 58, 57, 56
+    ]);
+  });
+
+  it('throws error with invalid input', () => {
+    // tslint:disable-next-line:no-any
+    const x: any = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
+    expect(() => tf.reverse4d(x, [1])).toThrowError();
+  });
+
+  it('throws error with invalid axis param', () => {
+    const x = tf.tensor4d([1, 20, 300, 4], [1, 1, 1, 4]);
+    expect(() => tf.reverse4d(x, [4])).toThrowError();
+    expect(() => tf.reverse4d(x, [-5])).toThrowError();
+  });
+
+  it('throws error with non integer axis param', () => {
+    const x = tf.tensor4d([1, 20, 300, 4], [1, 1, 1, 4]);
+    expect(() => tf.reverse4d(x, [0.5])).toThrowError();
+  });
+
+  it('accepts a tensor-like object', async () => {
+    const input = [[[[1]], [[2]], [[3]]], [[[4]], [[5]], [[6]]]];  // 2x3x1x1
+    const result = tf.reverse4d(input, [0]);
+    expect(result.shape).toEqual([2, 3, 1, 1]);
+    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
+  });
+});
diff --git a/tfjs-core/src/ops/reverse_test.ts b/tfjs-core/src/ops/reverse_test.ts
index 3ebf3351f49..ab377449a22 100644
--- a/tfjs-core/src/ops/reverse_test.ts
+++ b/tfjs-core/src/ops/reverse_test.ts
@@ -19,376 +19,6 @@ import * as tf from '../index';
 import {ALL_ENVS, describeWithFlags} from '../jasmine_util';
 import {expectArraysClose} from '../test_util';
 
-describeWithFlags('reverse1d', ALL_ENVS, () => {
-  it('reverse a 1D array', async () => {
-    const input = tf.tensor1d([1, 2, 3, 4, 5]);
-    const result = tf.reverse1d(input);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [5, 4, 3, 2, 1]);
-  });
-
-  it('reverse a 1D array, even length', async () => {
-    const input = tf.tensor1d([1, 2, 3, 4]);
-    const result = tf.reverse1d(input);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [4, 3, 2, 1]);
-  });
-
-  it('grad', async () => {
-    const a = tf.tensor1d([1, 2, 3]);
-    const dy = tf.tensor1d([10, 20, 30]);
-    const da = tf.grad((a: tf.Tensor1D) => tf.reverse1d(a))(a, dy);
-    expect(da.shape).toEqual([3]);
-    expectArraysClose(await da.data(), [30, 20, 10]);
-  });
-
-  it('gradient with clones', async () => {
-    const a = tf.tensor1d([1, 2, 3]);
-    const dy = tf.tensor1d([10, 20, 30]);
-    const da =
-        tf.grad((a: tf.Tensor1D) => tf.reverse1d(a.clone()).clone())(a, dy);
-    expect(da.shape).toEqual([3]);
-    expectArraysClose(await da.data(), [30, 20, 10]);
-  });
-
-  it('accepts a tensor-like object', async () => {
-    const input = [1, 2, 3, 4, 5];
-    const result = tf.reverse1d(input);
-    expect(result.shape).toEqual([5]);
-    expectArraysClose(await result.data(), [5, 4, 3, 2, 1]);
-  });
-});
-
-describeWithFlags('reverse2d', ALL_ENVS, () => {
-  it('reverse a 2D array at axis [0]', async () => {
-    const axis = [0];
-    const a = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
-    const result = tf.reverse2d(a, axis);
-
-    expect(result.shape).toEqual(a.shape);
-    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
-  });
-
-  it('reverse a 2D array at axis [1]', async () => {
-    const axis = [1];
-    const a = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
-    const result = tf.reverse2d(a, axis);
-
-    expect(result.shape).toEqual(a.shape);
-    expectArraysClose(await result.data(), [3, 2, 1, 6, 5, 4]);
-  });
-
-  it('reverse a 2D array odd rows and columns at axis [0, 1]', async () => {
-    const axis = [0, 1];
-    const a = tf.tensor2d(
-        [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], [3, 5]);
-    const result = tf.reverse2d(a, axis);
-
-    expect(result.shape).toEqual(a.shape);
-    expectArraysClose(
-        await result.data(),
-        [15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1]);
-  });
-
-  it('throws error with invalid input', () => {
-    // tslint:disable-next-line:no-any
-    const x: any = tf.tensor1d([1, 20, 300, 4]);
-    expect(() => tf.reverse2d(x, [0])).toThrowError();
-  });
-
-  it('throws error with invalid axis param', () => {
-    const x = tf.tensor2d([1, 20, 300, 4], [1, 4]);
-    expect(() => tf.reverse2d(x, [2])).toThrowError();
-    expect(() => tf.reverse2d(x, [-3])).toThrowError();
-  });
-
-  it('throws error with non integer axis param', () => {
-    const x = tf.tensor2d([1, 20, 300, 4], [1, 4]);
-    expect(() => tf.reverse2d(x, [0.5])).toThrowError();
-  });
-
-  it('grad', async () => {
-    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
-    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
-    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a))(a, dy);
-    expect(da.shape).toEqual([2, 3]);
-    expectArraysClose(await da.data(), [60, 50, 40, 30, 20, 10]);
-  });
-
-  it('grad with reverse(axis=0)', async () => {
-    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
-    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
-    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a, 0))(a, dy);
-    expect(da.shape).toEqual([2, 3]);
-    expectArraysClose(await da.data(), [40, 50, 60, 10, 20, 30]);
-  });
-
-  it('grad with reverse(axis=1)', async () => {
-    const a = tf.tensor2d([[1, 2, 3], [4, 5, 6]]);
-    const dy = tf.tensor2d([[10, 20, 30], [40, 50, 60]]);
-    const da = tf.grad((a: tf.Tensor2D) => tf.reverse2d(a, 1))(a, dy);
-    expect(da.shape).toEqual([2, 3]);
-    expectArraysClose(await da.data(), [30, 20, 10, 60, 50, 40]);
-  });
-
-  it('accepts a tensor-like object', async () => {
-    const axis = [0];
-    const a = [[1, 2, 3], [4, 5, 6]];  // 2x3
-    const result = tf.reverse2d(a, axis);
-
-    expect(result.shape).toEqual([2, 3]);
-    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
-  });
-});
-
-describeWithFlags('reverse3d', ALL_ENVS, () => {
-  // [
-  //   [
-  //     [0,  1,  2,  3],
-  //     [4,  5,  6,  7],
-  //     [8,  9,  10, 11]
-  //   ],
-  //   [
-  //     [12, 13, 14, 15],
-  //     [16, 17, 18, 19],
-  //     [20, 21, 22, 23]
-  //   ]
-  // ]
-  const shape: [number, number, number] = [2, 3, 4];
-  const data = [
-    0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11,
-    12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
-  ];
-
-  it('reverse a 3D array at axis [0]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [0]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
-      0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11
-    ]);
-  });
-
-  it('reverse a 3D array at axis [1]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [1]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3,
-      20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15
-    ]);
-  });
-
-  it('reverse a 3D array at axis [2]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [2]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8,
-      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20
-    ]);
-  });
-
-  it('reverse a 3D array at axis [0, 1]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [0, 1]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15,
-      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3
-    ]);
-  });
-
-  it('reverse a 3D array at axis [0, 2]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [0, 2]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20,
-      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8
-    ]);
-  });
-
-  it('reverse a 3D array at axis [1, 2]', async () => {
-    const input = tf.tensor3d(data, shape);
-    const result = tf.reverse3d(input, [1, 2]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      11, 10, 9,  8,  7,  6,  5,  4,  3,  2,  1,  0,
-      23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12
-    ]);
-  });
-
-  it('throws error with invalid input', () => {
-    // tslint:disable-next-line:no-any
-    const x: any = tf.tensor2d([1, 20, 300, 4], [1, 4]);
-    expect(() => tf.reverse3d(x, [1])).toThrowError();
-  });
-
-  it('throws error with invalid axis param', () => {
-    const x = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
-    expect(() => tf.reverse3d(x, [3])).toThrowError();
-    expect(() => tf.reverse3d(x, [-4])).toThrowError();
-  });
-
-  it('throws error with non integer axis param', () => {
-    const x = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
-    expect(() => tf.reverse3d(x, [0.5])).toThrowError();
-  });
-
-  it('accepts a tensor-like object', async () => {
-    const input = [[[1], [2], [3]], [[4], [5], [6]]];  // 2x3x1
-    const result = tf.reverse3d(input, [0]);
-    expect(result.shape).toEqual([2, 3, 1]);
-    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
-  });
-});
-
-describeWithFlags('reverse4d', ALL_ENVS, () => {
-  // [
-  //   [
-  //     [
-  //       [0,  1,  2,  3],
-  //       [4,  5,  6,  7],
-  //       [8,  9,  10, 11]
-  //     ],
-  //     [
-  //       [12, 13, 14, 15],
-  //       [16, 17, 18, 19],
-  //       [20, 21, 22, 23]
-  //     ]
-  //   ],
-  //   [
-  //     [
-  //       [24, 25, 26, 27],
-  //       [28, 29, 30, 31],
-  //       [32, 33, 34, 35]
-  //     ],
-  //     [
-  //       [36, 37, 38, 39],
-  //       [40, 41, 42, 43],
-  //       [44, 45, 46, 47]
-  //     ]
-  //   ],
-  //   [
-  //     [
-  //       [48, 49, 50, 51],
-  //       [52, 53, 54, 55],
-  //       [56, 57, 58, 59]
-  //     ],
-  //     [
-  //       [60, 61, 62, 63],
-  //       [64, 65, 66, 67],
-  //       [68, 69, 70, 71]
-  //     ]
-  //   ]
-  // ]
-  const shape: [number, number, number, number] = [3, 2, 3, 4];
-  const data = [
-    0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11, 12, 13, 14, 15, 16, 17,
-    18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
-    36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
-    54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71
-  ];
-
-  it('reverse a 4D array at axis [0]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [0]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65,
-      66, 67, 68, 69, 70, 71, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
-      36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 0,  1,  2,  3,  4,  5,
-      6,  7,  8,  9,  10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
-    ]);
-  });
-
-  it('reverse a 4D array at axis [1]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [1]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 0,  1,  2,  3,  4,  5,
-      6,  7,  8,  9,  10, 11, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47,
-      24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 60, 61, 62, 63, 64, 65,
-      66, 67, 68, 69, 70, 71, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59
-    ]);
-  });
-
-  it('reverse a 4D array at axis [2]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [2]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      8,  9,  10, 11, 4,  5,  6,  7,  0,  1,  2,  3,  20, 21, 22, 23, 16, 17,
-      18, 19, 12, 13, 14, 15, 32, 33, 34, 35, 28, 29, 30, 31, 24, 25, 26, 27,
-      44, 45, 46, 47, 40, 41, 42, 43, 36, 37, 38, 39, 56, 57, 58, 59, 52, 53,
-      54, 55, 48, 49, 50, 51, 68, 69, 70, 71, 64, 65, 66, 67, 60, 61, 62, 63
-    ]);
-  });
-
-  it('reverse a 4D array at axis [3]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [3]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      3,  2,  1,  0,  7,  6,  5,  4,  11, 10, 9,  8,  15, 14, 13, 12, 19, 18,
-      17, 16, 23, 22, 21, 20, 27, 26, 25, 24, 31, 30, 29, 28, 35, 34, 33, 32,
-      39, 38, 37, 36, 43, 42, 41, 40, 47, 46, 45, 44, 51, 50, 49, 48, 55, 54,
-      53, 52, 59, 58, 57, 56, 63, 62, 61, 60, 67, 66, 65, 64, 71, 70, 69, 68
-    ]);
-  });
-
-  it('reverse a 4D array at axis [0, 2]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [0, 2]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      56, 57, 58, 59, 52, 53, 54, 55, 48, 49, 50, 51, 68, 69, 70, 71, 64, 65,
-      66, 67, 60, 61, 62, 63, 32, 33, 34, 35, 28, 29, 30, 31, 24, 25, 26, 27,
-      44, 45, 46, 47, 40, 41, 42, 43, 36, 37, 38, 39, 8,  9,  10, 11, 4,  5,
-      6,  7,  0,  1,  2,  3,  20, 21, 22, 23, 16, 17, 18, 19, 12, 13, 14, 15
-    ]);
-  });
-
-  it('reverse a 4D array at axis [1, 3]', async () => {
-    const input = tf.tensor4d(data, shape);
-    const result = tf.reverse4d(input, [1, 3]);
-    expect(result.shape).toEqual(input.shape);
-    expectArraysClose(await result.data(), [
-      15, 14, 13, 12, 19, 18, 17, 16, 23, 22, 21, 20, 3,  2,  1,  0,  7,  6,
-      5,  4,  11, 10, 9,  8,  39, 38, 37, 36, 43, 42, 41, 40, 47, 46, 45, 44,
-      27, 26, 25, 24, 31, 30, 29, 28, 35, 34, 33, 32, 63, 62, 61, 60, 67, 66,
-      65, 64, 71, 70, 69, 68, 51, 50, 49, 48, 55, 54, 53, 52, 59, 58, 57, 56
-    ]);
-  });
-
-  it('throws error with invalid input', () => {
-    // tslint:disable-next-line:no-any
-    const x: any = tf.tensor3d([1, 20, 300, 4], [1, 1, 4]);
-    expect(() => tf.reverse4d(x, [1])).toThrowError();
-  });
-
-  it('throws error with invalid axis param', () => {
-    const x = tf.tensor4d([1, 20, 300, 4], [1, 1, 1, 4]);
-    expect(() => tf.reverse4d(x, [4])).toThrowError();
-    expect(() => tf.reverse4d(x, [-5])).toThrowError();
-  });
-
-  it('throws error with non integer axis param', () => {
-    const x = tf.tensor4d([1, 20, 300, 4], [1, 1, 1, 4]);
-    expect(() => tf.reverse4d(x, [0.5])).toThrowError();
-  });
-
-  it('accepts a tensor-like object', async () => {
-    const input = [[[[1]], [[2]], [[3]]], [[[4]], [[5]], [[6]]]];  // 2x3x1x1
-    const result = tf.reverse4d(input, [0]);
-    expect(result.shape).toEqual([2, 3, 1, 1]);
-    expectArraysClose(await result.data(), [4, 5, 6, 1, 2, 3]);
-  });
-});
-
 describeWithFlags('reverse', ALL_ENVS, () => {
   it('throws when passed a non-tensor', () => {
     expect(() => tf.reverse({} as tf.Tensor))
diff --git a/tfjs-core/src/public/chained_ops/register_all_chained_ops.ts b/tfjs-core/src/public/chained_ops/register_all_chained_ops.ts
index 5c71468725f..3267fbc4f58 100644
--- a/tfjs-core/src/public/chained_ops/register_all_chained_ops.ts
+++ b/tfjs-core/src/public/chained_ops/register_all_chained_ops.ts
@@ -62,6 +62,7 @@ import './relu';
 import './resize_bilinear';
 import './resize_nearest_neighbor';
 import './relu6';
+import './reverse';
 import './selu';
 import './separable_conv2d';
 import './split';
diff --git a/tfjs-core/src/public/chained_ops/register_all_chained_ops_test.ts b/tfjs-core/src/public/chained_ops/register_all_chained_ops_test.ts
index a35227e811c..33a984536ff 100644
--- a/tfjs-core/src/public/chained_ops/register_all_chained_ops_test.ts
+++ b/tfjs-core/src/public/chained_ops/register_all_chained_ops_test.ts
@@ -72,6 +72,7 @@ const CHAINED_OPS = [
   'resizeBilinear',
   'resizeNearestNeighbor',
   'relu6',
+  'reverse',
   'selu',
   'separableConv2d',
   'spaceToBatchND',
diff --git a/tfjs-core/src/public/chained_ops/reverse.ts b/tfjs-core/src/public/chained_ops/reverse.ts
new file mode 100644
index 00000000000..d81fa1acd68
--- /dev/null
+++ b/tfjs-core/src/public/chained_ops/reverse.ts
@@ -0,0 +1,31 @@
+/**
+ * @license
+ * Copyright 2020 Google LLC. All Rights Reserved.
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ * =============================================================================
+ */
+import {reverse} from '../../ops/reverse';
+import {Tensor} from '../../tensor';
+import {Rank} from '../../types';
+
+declare module '../../tensor' {
+  interface Tensor<R extends Rank = Rank> {
+    reverse<T extends Tensor>(this: T, axis?: number|number[]): T;
+  }
+}
+
+Tensor.prototype.reverse = function<T extends Tensor>(
+    this: T, axis?: number|number[]): T {
+  this.throwIfDisposed();
+  return reverse(this, axis);
+};
diff --git a/tfjs-core/src/register_all_gradients.ts b/tfjs-core/src/register_all_gradients.ts
index 563fa8d81d7..0bbf7c0d7ee 100644
--- a/tfjs-core/src/register_all_gradients.ts
+++ b/tfjs-core/src/register_all_gradients.ts
@@ -51,6 +51,7 @@ import {relu6GradConfig} from './gradients/Relu6_grad';
 import {reluGradConfig} from './gradients/Relu_grad';
 import {resizeBilinearGradConfig} from './gradients/ResizeBilinear_grad';
 import {resizeNearestNeighborGradConfig} from './gradients/ResizeNearestNeighbor_grad';
+import {reverseGradConfig} from './gradients/Reverse_grad';
 import {selectV2PoolGradConfig} from './gradients/SelectV2_grad';
 import {seluGradConfig} from './gradients/Selu_grad';
 import {spaceToBatchNDGradConfig} from './gradients/SpaceToBatchND_grad';
@@ -107,6 +108,7 @@ const gradConfigs: GradConfig[] = [
   resizeBilinearGradConfig,
   resizeNearestNeighborGradConfig,
   relu6GradConfig,
+  reverseGradConfig,
   seluGradConfig,
   selectV2PoolGradConfig,
   spaceToBatchNDGradConfig,
diff --git a/tfjs-core/src/tensor.ts b/tfjs-core/src/tensor.ts
index d0231ac3bf9..5dfe6e719d8 100644
--- a/tfjs-core/src/tensor.ts
+++ b/tfjs-core/src/tensor.ts
@@ -184,7 +184,6 @@ export interface OpHandler {
       keepDims: boolean): Tensor;
   slice<R extends Rank, T extends Tensor<R>>(
       x: T, begin: number|number[], size?: number|number[]): T;
-  reverse<T extends Tensor>(x: T, axis?: number|number[]): T;
   stack<T extends Tensor>(tensors: Array<T|TensorLike>, axis: number): Tensor;
   unstack<T extends Tensor>(value: T, axis: number): Tensor[];
   all<T extends Tensor>(x: Tensor, axis: number|number[], keepDims: boolean): T;
@@ -678,10 +677,6 @@ export class Tensor<R extends Rank = Rank> {
     this.throwIfDisposed();
     return opHandler.slice(this, begin, size);
   }
-  reverse<T extends Tensor>(this: T, axis?: number|number[]): T {
-    this.throwIfDisposed();
-    return opHandler.reverse(this, axis);
-  }
   stack(x: Tensor, axis = 0): Tensor {
     return opHandler.stack([this, x], axis);
   }
diff --git a/tfjs-core/src/tests.ts b/tfjs-core/src/tests.ts
index 3cc1ebe433a..e9a68fe0ac4 100644
--- a/tfjs-core/src/tests.ts
+++ b/tfjs-core/src/tests.ts
@@ -122,6 +122,10 @@ import './ops/relu6_test';
 import './ops/relu_test';
 import './ops/resize_bilinear_test';
 import './ops/resize_nearest_neighbor_test';
+import './ops/reverse_1d_test';
+import './ops/reverse_2d_test';
+import './ops/reverse_3d_test';
+import './ops/reverse_4d_test';
 import './ops/reverse_test';
 import './ops/scatter_nd_test';
 import './ops/segment_ops_test';