[bug]: Certificate error with patchmatch

[artatgray]

Is there an existing issue for this problem?

• I have searched the existing issues

Install method

Invoke's Launcher

Operating system

Windows

GPU vendor

Nvidia (CUDA)

GPU model

RTX 3080

GPU VRAM

16GB

Version number

InvokeAI v6.10.0

Browser

No response

System Information

Device name Bill-PC
Processor 13th Gen Intel(R) Core(TM) i9-13900H (2.60 GHz)
Installed RAM 64.0 GB (63.7 GB usable)
Device ID 3263F358-01C9-43BC-B821-0DEB2B3B5C0A
Product ID 00330-80000-00000-AA191
System type 64-bit operating system, x64-based processor
Pen and touch Pen support

Edition Windows 11 Pro
Version 25H2
Installed on ‎11/‎12/‎2024
OS build 26200.7623
Experience Windows Feature Experience Pack 1000.26100.275.0

What happened

When initiating from launcher v1.8.1 I receive the following error:

Preparing first run of this install - may take a minute or two...
Started Invoke process with PID 53500
[2026-01-30 19:56:52,408]::[InvokeAI]::INFO --> PyTorch CUDA memory allocator: cudaMallocAsync
[2026-01-30 19:56:57,457]::[InvokeAI]::INFO --> Using torch device: NVIDIA GeForce RTX 3080 Laptop GPU
[2026-01-30 19:56:58,493]::[InvokeAI]::INFO --> cuDNN version: 90701

patchmatch.patch_match: ERROR - patchmatch failed to load or compile (<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1010)>).
patchmatch.patch_match: INFO - Refer to https://invoke-ai.github.io/InvokeAI/installation/060_INSTALL_PATCHMATCH/ for installation instructions.
[2026-01-30 19:57:00,025]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal)

What you expected to happen

The docs state that patchmatch should be automatically installed on Windows. Even after a repair from the launcher the error still appears.

How to reproduce the problem

Preparing first run of this install - may take a minute or two...
Started Invoke process with PID 53500
[2026-01-30 19:56:52,408]::[InvokeAI]::INFO --> PyTorch CUDA memory allocator: cudaMallocAsync
[2026-01-30 19:56:57,457]::[InvokeAI]::INFO --> Using torch device: NVIDIA GeForce RTX 3080 Laptop GPU
[2026-01-30 19:56:58,493]::[InvokeAI]::INFO --> cuDNN version: 90701

patchmatch.patch_match: ERROR - patchmatch failed to load or compile (<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1010)>).
patchmatch.patch_match: INFO - Refer to https://invoke-ai.github.io/InvokeAI/installation/060_INSTALL_PATCHMATCH/ for installation instructions.
[2026-01-30 19:57:00,025]::[InvokeAI]::INFO --> Patchmatch not loaded (nonfatal)
[2026-01-30 19:57:00,743]::[InvokeAI]::INFO --> InvokeAI version 6.10.0
[2026-01-30 19:57:00,744]::[InvokeAI]::INFO --> Root directory = C:\Users\artat\invokeai
[2026-01-30 19:57:00,746]::[InvokeAI]::INFO --> Initializing database at C:\Users\artat\invokeai\databases\invokeai.db
[2026-01-30 19:57:00,823]::[InvokeAI]::INFO --> Database update needed
[2026-01-30 19:57:00,824]::[InvokeAI]::INFO --> Backing up database to C:\Users\artat\invokeai\databases\invokeai_backup_20260130-195700.db
[2026-01-30 19:57:01,009]::[InvokeAI]::INFO --> Database updated successfully
[2026-01-30 19:57:01,069]::[ModelManagerService]::INFO --> [MODEL CACHE] Calculated model RAM cache size: 13311.50 MB. Heuristics applied: [1, 2].
[2026-01-30 19:57:01,882]::[InvokeAI]::INFO --> Invoke running on http://127.0.0.1:9090 (Press CTRL+C to quit)

Additional context

No response

Discord username

No response

[lstein]

We may need to update our code signing certificate. I will investigate.

[lstein]

On further inspection, this appears to be a failure to download the patchmatch DLL from https://github.com/invoke-ai/PyPatchMatch/releases/download/0.1.1/libpatchmatch_windows_amd64.dll . This could be a transient network error, a firewall issue, or problem on your desktop computer. Could you cut and paste the URL into a browser running on your desktop and see if the DLL is downloadable? If it is, could you try the launcher again, assuming it was a transient network error?

[artatgray]

Hi,

The link was downloadable from my machine. However, when I tried the launcher again the same certificate verification error as initially reported occurred.

Thanks.

[lstein]

Do you have any sort of VPN or web proxy running on your machine/network?

[lstein]

I think you may have outdated X.509 certificates in your Python instance. Could you try this:

1. Start the launcher
2. Click on the ">" icon in the bottom left.
3. Click "Start dev console"
4. Type uv pip install --upgrade certifi and return

Restart the launcher and see if the install gets patchmatch successfully.

[artatgray]

Upon opening the dev console, I noticed the following:

Could that be causing the problem?

[lstein]

This is not directly related to the original problem, but has to be overcome in order to effect the solution. To get beyond this:

1. Open up a powershell on your system using the "As Administrator" version.
2. Type Set-ExecutionPolicy RemoteSigned
3. Try opening the launcher console again. The error message should be gone.
4. If you still get the error, try Set-ExecutionPolicy Unrestricted

Assuming the certifi update works, you should put the security restriction back the way it was with Set-ExecutionPolicy Restricted

[artatgray]

Regarding setting policies, I've ever done everything that you've suggested and it's still throwing up the error.

I decided to investigate a bit via the dev console and the following is what I found. Perhaps it will offer more insight:

[github-actions]

There has been no activity in this issue for 30 days. If this issue is still being experienced, please reply with an updated confirmation that the issue is still being experienced with the latest release.

[github-actions]

Due to inactivity, this issue was automatically closed. If you are still experiencing the issue, please recreate the issue.