Interactive Python Window only Works Intermittently

[Nick-Hemenway]

Type: Bug

My colleague and I are both running into an issue where when we try to run a python script in an interactive window, the terminal only starts up and runs intermittently.

I don't know how to repeatably replicate this as it seems completely sporadic, but the actions that produce it are:
Right click in editor with python script open --> Run in Interactive Window --> Run Current File in Interactive Window

Behavior: Sometimes the interactive window will open and run the code, but often the interactive window will open and appears to be stuck trying to connect to the python terminal. It will just remain frozen indefinitely. I then kill the interactive terminal and repeatedly try to run the python script in the interactive terminal until eventually it either runs, or I get a window popping up saying "Unknown Text Editor Error", at which point I have to restart VS-Code.

As you can see in the video below, I have a simple script that I repeatedly try to run. The first three times the interactive window opens but it remains in a frozen state, unable to actually run the code. Finally, the fourth time the code runs. We've been experiencing this super frustrating bug for the past year.

jupyter_extension_bug.mp4

Extension version: 2025.9.1
VS Code version: Code 1.115.0 (41dd792b5e652393e7787322889ed5fdc58bd75b, 2026-04-07T22:04:57Z)
OS version: Windows_NT x64 10.0.26200
Modes:

System Info

Item	Value
CPUs	13th Gen Intel(R) Core(TM) i9-13950HX (32 x 2419)
GPU Status	2d_canvas: enabled GPU0: VENDOR= 0x10de, DEVICE=0x2730 [NVIDIA RTX 5000 Ada Generation Laptop GPU], DRIVER_VENDOR=NVIDIA, DRIVER_VERSION=32.0.15.8160 ACTIVE GPU1: VENDOR= 0x8086, DEVICE=0xa788 [Intel(R) UHD Graphics], DRIVER_VERSION=32.0.101.7082 GPU2: VENDOR= 0x1414, DEVICE=0x008c [Microsoft Basic Render Driver], DRIVER_VERSION=10.0.26100.7309 Machine model name: Machine model version: direct_rendering_display_compositor: disabled_off_ok gpu_compositing: enabled multiple_raster_threads: enabled_on opengl: enabled_on rasterization: enabled raw_draw: disabled_off_ok skia_graphite: disabled_off trees_in_viz: disabled_off video_decode: enabled video_encode: enabled webgl: enabled webgl2: enabled webgpu: enabled webnn: disabled_off
Load (avg)	undefined
Memory (System)	127.66GB (84.17GB free)
Process Argv	--crash-reporter-id 08eb9a1d-4d28-4089-8f22-0cfc40e29f5a
Screen Reader	no
VM	0%

A/B Experiments

vsliv368:30146709
binariesv615:30325510
nativeloc1:31344060
dwcopilot:31170013
dwoutputs:31242946
copilot_t_ci:31333650
e5gg6876:31282496
pythonrdcb7:31342333
6518g693:31463988
aj953862:31281341
6abeh943:31336334
envsactivate1:31464700
cloudbuttont:31379625
3efgi100_wstrepl:31403338
use-responses-api:31390855
anonusage:31403420
je187915:31401257
ec5jj548:31422691
cp_cls_c_966_ss:31454199
inlinechat_v2_hd992725:31485760
4je02754:31466945
8hhj4413:31478653
ge8j1254_inline_auto_hint_haiku:31490510
38bie571_auto:31478677
cp_cls_c_1081:31454833
conptydll_true:31489736
e9c30283:31461165
test_control1:31471004
idci7584:31464702
e3e4d672:31489863
534a6447_copy:31478748
hg17d649:31458077
chat:31457767
t-some:31486549
cpptoolsoff-v2:31475362
inline_hover_fd9bg283:31492073
ec1eg703:31491645
ddid_c:31478207
getcmakediagnosticson:31489824
pro_large_t:31492544
cp_cls_c_1082:31491634
73afh133:31491943
logging_disabled_new:31490726

[carlosanapo]

Same for me. I noticed this actually only happens when a virtual environment is used (even if the interpreter version and kernel are fully compatible and "brand new").

[Nick-Hemenway]

Same for me. I noticed this actually only happens when a virtual environment is used (even if the interpreter version and kernel are fully compatible and "brand new").

Since switching to using uv for all of my python packaging needs, I exclusively use virtual environments and can't comment on whether this issue goes away when using a global python installation. How do you have your python environment set up? I'd rather not use global python installations as a workaround.

[carlosanapo]

Same for me. I noticed this actually only happens when a virtual environment is used (even if the interpreter version and kernel are fully compatible and "brand new").

Since switching to using uv for all of my python packaging needs, I exclusively use virtual environments and can't comment on whether this issue goes away when using a global python installation. How do you have your python environment set up? I'd rather not use global python installations as a workaround.

Indeed, I do not like switching to the global python too. But sometimes is the only solution. Regarding the venv I generate it via command line in a terminal, then I register the corresponding ipykernel to be used in the interactive window. That's it, I guess.

[amunger]

does anything similar happen when trying to use that same kernel with a notebook, or is it specific to the interactive window?
And does it only occur with the Run Current File in Interactive Window command?

Can you add "jupyter.logging.level": "trace" to your settings, restart vscode, then repro the issue and copy the jupyter logs from the output window and attach those to the issue.