[GH-ISSUE #17376] issue: Disabling CPU for Ollama removes all models from open-webui #72864

New Issue

GiteaMirror · 2026-05-13T05:03:39-05:00

GiteaMirror commented

2026-05-13 05:03:39 -05:00

Originally created by @ryanjoneil on GitHub (Sep 11, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/17376

Check Existing Issues

I have searched for any existing and/or related issues.
I have searched for any existing and/or related discussions.
I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.6.28

Ollama Version (if applicable)

0.11.10

Operating System

Ubuntu 25.04

Browser (if applicable)

Firefox 142.0.1

Confirmation

I have read and followed all instructions in README.md.
I am using the latest version of both Open WebUI and Ollama.
I have included the browser console logs.
I have included the Docker container logs.
I have provided every relevant configuration, setting, and environment variable used in my setup.
I have clearly listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
I have documented step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation. My steps:
Start with the initial platform/version/OS and dependencies used,
Specify exact install/launch/configure commands,
List URLs visited, user input (incl. example values/emails/passwords if needed),
Describe all options and toggles enabled or changed,
Include any files or environmental changes,
Identify the expected and actual result at each stage,
Ensure any reasonably skilled user can follow and hit the same issue.

Expected Behavior

When I configure Ollama to disable the CPU, the list of available models should still show up in open-webui.

Actual Behavior

When I configure Ollama to disable the CPU, the list of models in open-webui is empty. When I re-enable the CPU, the list of models in open-webui returns.

Steps to Reproduce

Install Ollama using their curl command.

https://ollama.com/download

ollama pull a number of different models.
Install and start open-webui using docker.

docker pull ghcr.io/open-webui/open-webui:main
docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main

Load any of the pages in open-webui with a model list (e.g. /admin/settings/models or /). They will show a list of available models, such as gpt-oss:20b.
Disable CPU for Ollama and restart the service.

sudo systemctl edit ollama

[Service]
Environment="OLLAMA_DISABLE_CPU=1 OLLAMA_HOST=0.0.0.0:11434"

sudo systemctl restart ollama

Reload the pages, and the model list is empty. However, in both situations the /api/v1/configs/models endpoint returns this data for me, which indicates that something funny is happening in the user interface. It's able to see all the models in Ollama, it just isn't showing them to me.

{
  "DEFAULT_MODELS": "gpt-oss:20b",
  "MODEL_ORDER_LIST": [
    "arena-model",
    "deepseek-r1:14b",
    "deepseek-r1:32b",
    "deepseek-r1:8b",
    "gemma3:27b",
    "gpt-oss:120b",
    "gpt-oss:20b"
  ]
}

Logs & Screenshots

Screenshots

Models available with CPU enabled

Models available with CPU disabled

Logs

Additional Information

No response

Originally created by @ryanjoneil on GitHub (Sep 11, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/17376 ### Check Existing Issues - [x] I have searched for any existing and/or related issues. - [x] I have searched for any existing and/or related discussions. - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version v0.6.28 ### Ollama Version (if applicable) 0.11.10 ### Operating System Ubuntu 25.04 ### Browser (if applicable) Firefox 142.0.1 ### Confirmation - [x] I have read and followed all instructions in `README.md`. - [x] I am using the latest version of **both** Open WebUI and Ollama. - [x] I have included the browser console logs. - [x] I have included the Docker container logs. - [x] I have **provided every relevant configuration, setting, and environment variable used in my setup.** - [x] I have clearly **listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup** (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc). - [x] I have documented **step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation**. My steps: - Start with the initial platform/version/OS and dependencies used, - Specify exact install/launch/configure commands, - List URLs visited, user input (incl. example values/emails/passwords if needed), - Describe all options and toggles enabled or changed, - Include any files or environmental changes, - Identify the expected and actual result at each stage, - Ensure any reasonably skilled user can follow and hit the same issue. ### Expected Behavior When I configure Ollama to disable the CPU, the list of available models should still show up in open-webui. ### Actual Behavior When I configure Ollama to disable the CPU, the list of models in open-webui is empty. When I re-enable the CPU, the list of models in open-webui returns. ### Steps to Reproduce 1. Install Ollama using [their curl command](https://ollama.com/download). ```bash https://ollama.com/download ``` 2. `ollama pull` a number of different models. 3. Install and start open-webui using docker. ```bash docker pull ghcr.io/open-webui/open-webui:main docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main ``` 4. Load any of the pages in open-webui with a model list (e.g. `/admin/settings/models` or `/`). They will show a list of available models, such as `gpt-oss:20b`. 5. Disable CPU for Ollama and restart the service. ```bash sudo systemctl edit ollama ``` ```config [Service] Environment="OLLAMA_DISABLE_CPU=1 OLLAMA_HOST=0.0.0.0:11434" ``` ```bash sudo systemctl restart ollama ``` 6. Reload the pages, and the model list is empty. However, in both situations the `/api/v1/configs/models` endpoint returns this data for me, which indicates that something funny is happening in the user interface. It's able to see all the models in Ollama, it just isn't showing them to me. ```json { "DEFAULT_MODELS": "gpt-oss:20b", "MODEL_ORDER_LIST": [ "arena-model", "deepseek-r1:14b", "deepseek-r1:32b", "deepseek-r1:8b", "gemma3:27b", "gpt-oss:120b", "gpt-oss:20b" ] } ``` ### Logs & Screenshots ### Screenshots #### Models available with CPU enabled <img width="1320" height="898" alt="Image" src="https://github.com/user-attachments/assets/0e3d22fc-fc8d-4e71-a538-bd4485749922" /> #### Models available with CPU disabled <img width="1320" height="898" alt="Image" src="https://github.com/user-attachments/assets/c321dac7-223e-4975-a5a1-73e2fb80f34b" /> ### Logs * [open-webui.log](https://github.com/user-attachments/files/22281869/open-webui.log) * [ollama.log](https://github.com/user-attachments/files/22281868/ollama.log) ### Additional Information _No response_

GiteaMirror added the bug label 2026-05-13 05:03:39 -05:00

GiteaMirror closed this issue

2026-05-13 05:03:40 -05:00

Sign in to join this conversation.

Branches Tags

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/open-webui#72864