diff --git a/src/transformers/pipelines/image_text_to_text.py b/src/transformers/pipelines/image_text_to_text.py
index 6937b63ae6a8b4..f2a10208345768 100644
--- a/src/transformers/pipelines/image_text_to_text.py
+++ b/src/transformers/pipelines/image_text_to_text.py
@@ -208,14 +208,6 @@ def _sanitize_parameters(
                     " please use only one"
                 )
             forward_kwargs["generate_kwargs"]["max_new_tokens"] = max_new_tokens
-        else:
-            if "generate_kwargs" not in forward_kwargs:
-                forward_kwargs["generate_kwargs"] = {}
-            if "max_new_tokens" not in forward_kwargs["generate_kwargs"]:
-                logger.warning_once(
-                    "The `max_new_tokens` parameter is not set. By default, the model will generate up to 20 tokens."
-                )
-                forward_kwargs["generate_kwargs"]["max_new_tokens"] = 20
 
         if return_full_text is not None and return_type is None:
             if return_tensors is not None:
diff --git a/tests/pipelines/test_pipelines_image_text_to_text.py b/tests/pipelines/test_pipelines_image_text_to_text.py
index 9ededa28d20f94..d7f7c74bf3b061 100644
--- a/tests/pipelines/test_pipelines_image_text_to_text.py
+++ b/tests/pipelines/test_pipelines_image_text_to_text.py
@@ -71,7 +71,7 @@ def test_small_model_pt_token(self):
         image = "./tests/fixtures/tests_samples/COCO/000000039769.png"
         text = "<image> What this is? Assistant: This is"
 
-        outputs = pipe(image, text=text, max_new_tokens=20)
+        outputs = pipe(image, text=text)
         self.assertEqual(
             outputs,
             [
@@ -82,7 +82,7 @@ def test_small_model_pt_token(self):
             ],
         )
 
-        outputs = pipe([image, image], text=[text, text], max_new_tokens=20)
+        outputs = pipe([image, image], text=[text, text])
         self.assertEqual(
             outputs,
             [
@@ -103,8 +103,8 @@ def test_consistent_batching_behaviour(self):
         image = "./tests/fixtures/tests_samples/COCO/000000039769.png"
         prompt = "a photo of"
 
-        outputs = pipe([image, image], text=[prompt, prompt], max_new_tokens=20)
-        outputs_batched = pipe([image, image], text=[prompt, prompt], max_new_tokens=20, batch_size=2)
+        outputs = pipe([image, image], text=[prompt, prompt])
+        outputs_batched = pipe([image, image], text=[prompt, prompt], batch_size=2)
         self.assertEqual(outputs, outputs_batched)
 
     @slow
@@ -123,7 +123,7 @@ def test_model_pt_chat_template(self):
                 ],
             }
         ]
-        outputs = pipe([image_ny, image_chicago], text=messages, max_new_tokens=20)
+        outputs = pipe([image_ny, image_chicago], text=messages)
         self.assertEqual(
             outputs,
             [
@@ -178,7 +178,7 @@ def test_model_pt_chat_template_continue_final_message(self):
                 ],
             },
         ]
-        outputs = pipe(text=messages, max_new_tokens=20)
+        outputs = pipe(text=messages)
         self.assertEqual(
             outputs,
             [
@@ -237,7 +237,7 @@ def test_model_pt_chat_template_new_text(self):
                 ],
             }
         ]
-        outputs = pipe(text=messages, max_new_tokens=20, return_full_text=False)
+        outputs = pipe(text=messages, return_full_text=False)
         self.assertEqual(
             outputs,
             [