Enable RunCell server extension (JupyterLab restart required)
Fix local backend not running (local ping failed) by restarting JupyterLab and enabling the RunCell Jupyter Server extension when needed
When you need this
You might see RunCell installed in JupyterLab, but it doesn’t work or shows a setup/connection issue.
This usually means the RunCell Jupyter Server extension (the backend handler) is not running yet.
The most common cause: you installed RunCell via JupyterLab Plugin Manager, and JupyterLab needs a full restart (not a browser refresh) for the server-side part to load.
Step 1 — Fully restart JupyterLab (not refresh)
- Close the JupyterLab browser tab.
- Stop the running JupyterLab process in your terminal (usually
Ctrl + C). - Start it again:
jupyter labIf you installed via Plugin Manager, this step alone fixes most cases.
Step 2 — Verify the server extension is enabled
Run:
jupyter server extension listLook for something like runcell in the output.
If you’re using a virtualenv/conda environment, make sure you run the command in the same environment as jupyter lab.
Quick sanity check:
which jupyter
python -c "import sys; print(sys.executable)"
python -m pip show runcellStep 3 — Enable the server extension (only if needed)
If the extension is missing/disabled, enable it and restart JupyterLab.
Try:
jupyter server extension enable runcellIf that doesn’t work (common with conda/brew edge cases), try one of these variants and then restart:
# Enable for the current environment (recommended for venv/conda)
jupyter server extension enable --sys-prefix runcell
# Enable only for the current OS user
jupyter server extension enable --user runcellThen restart:
jupyter labCommon edge cases (conda / Homebrew / multiple Pythons)
- Environment mismatch:
jupyter labis running from one Python env, butruncellis installed in another.- Fix by installing RunCell into the same environment that provides
jupyter.
- Fix by installing RunCell into the same environment that provides
- Homebrew / system-wide Jupyter: you may need to ensure your terminal PATH points to the expected
jupyter. - Still stuck?
- Restart JupyterLab with debug logs:
jupyter lab --log-level=DEBUGIf you still can’t get it working, collect the output of:
python --version
jupyter lab --version
which python
which jupyter
jupyter server extension list
python -m pip show runcelland contact support.