huggingface · Rocketknight1 · Jan 16, 2026 · Jan 16, 2026 · Jan 16, 2026 · Jan 16, 2026
diff --git a/MIGRATION_GUIDE_V5.md b/MIGRATION_GUIDE_V5.md
@@ -609,9 +609,11 @@ Linked PRs:
 
 ## Pipelines
 
-`Text2TextGenerationPipeline`, as well as the related `SummarizationPipeline` and `TranslationPipeline`, were deprecated and will now be removed.
-`pipeline` classes are intended as a high-level beginner-friendly API, but for almost all text-to-text tasks a modern chat model 
-and `TextGenerationPipeline` will provide much higher quality output. As a result, we felt it was misleading for beginners to offer the older pipelines.
+### Text pipelines that should just be LLMs
+
+`question-answering` and `Text2TextGenerationPipeline`, including its related `SummarizationPipeline` and `TranslationPipeline`, were deprecated and will now be removed. `pipeline` classes are intended as a high-level beginner-friendly API,
+but for almost all text-to-text or question-answering tasks a modern chat model and `TextGenerationPipeline` will provide much higher quality output. 
+As a result, we felt it was misleading for beginners to offer the older pipelines.
 
 If you were using these pipelines before, try using `TextGenerationPipeline` with a chat model instead. For example, for summarization:
 
@@ -630,9 +632,12 @@ message_history = [
 print(summarizer(message_history)[0]["generated_text"][-1]["content"])
 ```
 
-Similarly, the `image-to-text` pipeline has been removed. This pipeline was used for early image captioning models, but these
-no longer offer competitive performance. Instead, for image captioning tasks we recommend using a modern vision-language chat model
-via the `image-text-to-text` pipeline. For example:
+The above example can be adapted for other tasks, e.g. translation or question answering, simply by changing the prompt.
+
+### Vision pipelines that should just be VLMs
+
+Similarly, the `image-to-text` and `visual-question-answering` pipelines have been removed. For image captioning or question answering
+tasks we recommend using a modern vision-language chat model via the `image-text-to-text` pipeline. For example:
 
 ```python
 import torch
@@ -655,6 +660,16 @@ message_history = [
 print(captioner(message_history)[0]["generated_text"][-1]["content"])
 ```
 
+The above example can be adapted for visual question answering simply by asking the question in the prompt.
+
+### Other removed pipelines
+
+The `image-to-image` pipeline has been removed, as it was rarely updated or used. For most image generation tasks, you
+probably want [🤗 Diffusers](https://huggingface.co/docs/diffusers/index) instead!
+
+### Other changes
+
+- The `feature-extraction` pipeline has now been renamed to `text-embedding` and the `image-feature-extraction` pipeline has been renamed to `image-embedding`. The older names are still usable as aliases, so this should not impact your existing code.
 - Image text to text pipelines will no longer accept images as a separate argument along with conversation chats. Image data has to be embedded in the chat's "content" field. See [#42359](https://github.com/huggingface/transformers/pull/42359)
 
 ## PushToHubMixin

diff --git a/docs/source/en/internal/pipelines_utils.md b/docs/source/en/internal/pipelines_utils.md
@@ -26,8 +26,6 @@ Most of those are only useful if you are studying the code of the models in the
 
 [[autodoc]] pipelines.ZeroShotClassificationArgumentHandler
 
-[[autodoc]] pipelines.QuestionAnsweringArgumentHandler
-
 ## Data format
 
 [[autodoc]] pipelines.PipelineDataFormat

diff --git a/docs/source/en/main_classes/pipelines.md b/docs/source/en/main_classes/pipelines.md
@@ -398,12 +398,6 @@ Pipelines available for natural language processing tasks include the following.
     - __call__
     - all
 
-### QuestionAnsweringPipeline
-
-[[autodoc]] QuestionAnsweringPipeline
-    - __call__
-    - all
-
 ### TableQuestionAnsweringPipeline
 
 [[autodoc]] TableQuestionAnsweringPipeline
@@ -443,15 +437,15 @@ Pipelines available for multimodal tasks include the following.
     - __call__
     - all
 
-### FeatureExtractionPipeline
+### TextEmbeddingPipeline
 
-[[autodoc]] FeatureExtractionPipeline
+[[autodoc]] TextEmbeddingPipeline
     - __call__
     - all
 
-### ImageFeatureExtractionPipeline
+### ImageEmbeddingPipeline
 
-[[autodoc]] ImageFeatureExtractionPipeline
+[[autodoc]] ImageEmbeddingPipeline
     - __call__
     - all
 
@@ -473,12 +467,6 @@ Pipelines available for multimodal tasks include the following.
     - __call__
     - all
 
-### VisualQuestionAnsweringPipeline
-
-[[autodoc]] VisualQuestionAnsweringPipeline
-    - __call__
-    - all
-
 ## Parent class: `Pipeline`
 
 [[autodoc]] Pipeline
diff --git a/docs/source/en/pipeline_tutorial.md b/docs/source/en/pipeline_tutorial.md
@@ -22,7 +22,7 @@ Tailor the [`Pipeline`] to your task with task specific parameters such as addin
 
 <Youtube id=tiZFewofSLM/>
 
-Transformers has two pipeline classes, a generic [`Pipeline`] and many individual task-specific pipelines like [`TextGenerationPipeline`] or [`VisualQuestionAnsweringPipeline`]. Load these individual pipelines by setting the task identifier in the `task` parameter in [`Pipeline`]. You can find the task identifier for each pipeline in their API documentation.
+Transformers has two pipeline classes, a generic [`Pipeline`] and many individual task-specific pipelines like [`TextGenerationPipeline`]. Load these individual pipelines by setting the task identifier in the `task` parameter in [`Pipeline`]. You can find the task identifier for each pipeline in their API documentation.
 
 Each task is configured to use a default pretrained model and preprocessor, but this can be overridden with the `model` parameter if you want to use a different model.
 

diff --git a/docs/source/ja/internal/pipelines_utils.md b/docs/source/ja/internal/pipelines_utils.md
@@ -27,8 +27,6 @@ rendered properly in your Markdown viewer.
 
 [[autodoc]] pipelines.ZeroShotClassificationArgumentHandler
 
-[[autodoc]] pipelines.QuestionAnsweringArgumentHandler
-
 ## Data format
 
 [[autodoc]] pipelines.PipelineDataFormat

diff --git a/docs/source/ja/main_classes/pipelines.md b/docs/source/ja/main_classes/pipelines.md
@@ -400,12 +400,6 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
 
 詳細については、[`TokenClassificationPipeline`] を参照してください。
 
-### QuestionAnsweringPipeline
-
-[[autodoc]] QuestionAnsweringPipeline
-    - __call__
-    - all
-
 ### TableQuestionAnsweringPipeline
 
 [[autodoc]] TableQuestionAnsweringPipeline
@@ -445,15 +439,15 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
     - __call__
     - all
 
-### FeatureExtractionPipeline
+### TextEmbeddingPipeline
 
-[[autodoc]] FeatureExtractionPipeline
+[[autodoc]] TextEmbeddingPipeline
     - __call__
     - all
 
-### ImageFeatureExtractionPipeline
+### ImageEmbeddingPipeline
 
-[[autodoc]] ImageFeatureExtractionPipeline
+[[autodoc]] ImageEmbeddingPipeline
     - __call__
     - all
 
@@ -463,12 +457,6 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
     - __call__
     - all
 
-### VisualQuestionAnsweringPipeline
-
-[[autodoc]] VisualQuestionAnsweringPipeline
-    - __call__
-    - all
-
 ## Parent class: `Pipeline`
 
 [[autodoc]] Pipeline
diff --git a/docs/source/ko/internal/pipelines_utils.md b/docs/source/ko/internal/pipelines_utils.md
@@ -26,8 +26,6 @@ rendered properly in your Markdown viewer.
 
 [[autodoc]] pipelines.ZeroShotClassificationArgumentHandler
 
-[[autodoc]] pipelines.QuestionAnsweringArgumentHandler
-
 ## 데이터 형식 [[transformers.PipelineDataFormat]]
 
 [[autodoc]] pipelines.PipelineDataFormat

diff --git a/docs/source/ko/main_classes/pipelines.md b/docs/source/ko/main_classes/pipelines.md
@@ -365,12 +365,6 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
     - __call__
     - all
 
-### QuestionAnsweringPipeline [[transformers.QuestionAnsweringPipeline]]
-
-[[autodoc]] QuestionAnsweringPipeline
-    - __call__
-    - all
-
 ### TableQuestionAnsweringPipeline [[transformers.TableQuestionAnsweringPipeline]]
 
 [[autodoc]] TableQuestionAnsweringPipeline
@@ -410,15 +404,15 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
     - __call__
     - all
 
-### FeatureExtractionPipeline [[transformers.FeatureExtractionPipeline]]
+### TextEmbeddingPipeline [[transformers.TextEmbeddingPipeline]]
 
-[[autodoc]] FeatureExtractionPipeline
+[[autodoc]] TextEmbeddingPipeline
     - __call__
     - all
 
-### ImageFeatureExtractionPipeline [[transformers.ImageFeatureExtractionPipeline]]
+### ImageEmbeddingPipeline [[transformers.ImageEmbeddingPipeline]]
 
-[[autodoc]] ImageFeatureExtractionPipeline
+[[autodoc]] ImageEmbeddingPipeline
     - __call__
     - all
 
@@ -434,12 +428,6 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
     - __call__
     - all
 
-### VisualQuestionAnsweringPipeline [[transformers.VisualQuestionAnsweringPipeline]]
-
-[[autodoc]] VisualQuestionAnsweringPipeline
-    - __call__
-    - all
-
 ## Parent class: `Pipeline` [[transformers.Pipeline]]
 
 [[autodoc]] Pipeline
diff --git a/docs/source/zh/internal/pipelines_utils.md b/docs/source/zh/internal/pipelines_utils.md
@@ -28,8 +28,6 @@ rendered properly in your Markdown viewer.
 
 [[autodoc]] pipelines.ZeroShotClassificationArgumentHandler
 
-[[autodoc]] pipelines.QuestionAnsweringArgumentHandler
-
 ## 数据格式
 
 [[autodoc]] pipelines.PipelineDataFormat

diff --git a/docs/source/zh/main_classes/pipelines.md b/docs/source/zh/main_classes/pipelines.md
@@ -374,12 +374,6 @@ my_pipeline = pipeline(model="xxxx", pipeline_class=MyPipeline)
 
 See [`TokenClassificationPipeline`] for all details.
 
-### QuestionAnsweringPipeline
-
-[[autodoc]] QuestionAnsweringPipeline
-    - __call__
-    - all
-
 ### TableQuestionAnsweringPipeline
 
 [[autodoc]] TableQuestionAnsweringPipeline
@@ -419,15 +413,15 @@ See [`TokenClassificationPipeline`] for all details.
     - __call__
     - all
 
-### FeatureExtractionPipeline
+### TextEmbeddingPipeline
 
-[[autodoc]] FeatureExtractionPipeline
+[[autodoc]] TextEmbeddingPipeline
     - __call__
     - all
 
-### ImageFeatureExtractionPipeline
+### ImageEmbeddingPipeline
 
-[[autodoc]] ImageFeatureExtractionPipeline
+[[autodoc]] ImageEmbeddingPipeline
     - __call__
     - all
 
@@ -443,12 +437,6 @@ See [`TokenClassificationPipeline`] for all details.
     - __call__
     - all
 
-### VisualQuestionAnsweringPipeline
-
-[[autodoc]] VisualQuestionAnsweringPipeline
-    - __call__
-    - all
-
 ## Parent class: `Pipeline`
 
 [[autodoc]] Pipeline
diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py
@@ -140,10 +140,9 @@
         "CsvPipelineDataFormat",
         "DepthEstimationPipeline",
         "DocumentQuestionAnsweringPipeline",
-        "FeatureExtractionPipeline",
         "FillMaskPipeline",
         "ImageClassificationPipeline",
-        "ImageFeatureExtractionPipeline",
+        "ImageEmbeddingPipeline",
         "ImageSegmentationPipeline",
         "ImageTextToTextPipeline",
         "ImageToImagePipeline",
@@ -155,14 +154,13 @@
         "PipedPipelineDataFormat",
         "Pipeline",
         "PipelineDataFormat",
-        "QuestionAnsweringPipeline",
         "TableQuestionAnsweringPipeline",
         "TextClassificationPipeline",
+        "TextEmbeddingPipeline",
         "TextGenerationPipeline",
         "TextToAudioPipeline",
         "TokenClassificationPipeline",
         "VideoClassificationPipeline",
-        "VisualQuestionAnsweringPipeline",
         "ZeroShotAudioClassificationPipeline",
         "ZeroShotClassificationPipeline",
         "ZeroShotImageClassificationPipeline",
@@ -648,10 +646,9 @@
     from .pipelines import CsvPipelineDataFormat as CsvPipelineDataFormat
     from .pipelines import DepthEstimationPipeline as DepthEstimationPipeline
     from .pipelines import DocumentQuestionAnsweringPipeline as DocumentQuestionAnsweringPipeline
-    from .pipelines import FeatureExtractionPipeline as FeatureExtractionPipeline
     from .pipelines import FillMaskPipeline as FillMaskPipeline
     from .pipelines import ImageClassificationPipeline as ImageClassificationPipeline
-    from .pipelines import ImageFeatureExtractionPipeline as ImageFeatureExtractionPipeline
+    from .pipelines import ImageEmbeddingPipeline as ImageEmbeddingPipeline
     from .pipelines import ImageSegmentationPipeline as ImageSegmentationPipeline
     from .pipelines import ImageTextToTextPipeline as ImageTextToTextPipeline
     from .pipelines import ImageToImagePipeline as ImageToImagePipeline
@@ -663,14 +660,13 @@
     from .pipelines import PipedPipelineDataFormat as PipedPipelineDataFormat
     from .pipelines import Pipeline as Pipeline
     from .pipelines import PipelineDataFormat as PipelineDataFormat
-    from .pipelines import QuestionAnsweringPipeline as QuestionAnsweringPipeline
     from .pipelines import TableQuestionAnsweringPipeline as TableQuestionAnsweringPipeline
     from .pipelines import TextClassificationPipeline as TextClassificationPipeline
+    from .pipelines import TextEmbeddingPipeline as TextEmbeddingPipeline
     from .pipelines import TextGenerationPipeline as TextGenerationPipeline
     from .pipelines import TextToAudioPipeline as TextToAudioPipeline
     from .pipelines import TokenClassificationPipeline as TokenClassificationPipeline
     from .pipelines import VideoClassificationPipeline as VideoClassificationPipeline
-    from .pipelines import VisualQuestionAnsweringPipeline as VisualQuestionAnsweringPipeline
     from .pipelines import ZeroShotAudioClassificationPipeline as ZeroShotAudioClassificationPipeline
     from .pipelines import ZeroShotClassificationPipeline as ZeroShotClassificationPipeline
     from .pipelines import ZeroShotImageClassificationPipeline as ZeroShotImageClassificationPipeline