diff --git a/.github/workflows/docs.yml b/.github/workflows/docs.yml
index 033f41de9fc..aa15bd67165 100644
--- a/.github/workflows/docs.yml
+++ b/.github/workflows/docs.yml
@@ -74,6 +74,9 @@ jobs:
         cd ./docs
         conda run -n build_binary make html
         cd ..
+    - name: Pull TensorDict docs
+      run: |
+        git subtree add --prefix=docs/build/html/tensordict --squash https://github.com/pytorch-labs/tensordict.git gh-pages
     - name: Get output time
       run: echo "The time was ${{ steps.build.outputs.time }}"
     - name: Deploy
diff --git a/docs/source/index.rst b/docs/source/index.rst
index 2eecbe8cf1a..39c90e1841b 100644
--- a/docs/source/index.rst
+++ b/docs/source/index.rst
@@ -6,6 +6,9 @@
 Welcome to the TorchRL Documentation!
 =====================================
 
+.. note::
+   The TensorDict class has been moved out of TorchRL into a dedicated library. Take a look at `the documentation <./tensordict>`_ or find the source code `on GitHub <https://github.com/pytorch-labs/tensordict>`_.
+
 TorchRL is an open-source Reinforcement Learning (RL) library for PyTorch.
 
 It provides pytorch and python-first, low and high level abstractions for RL that are intended to be efficient, modular, documented and properly tested.