[GH-ISSUE #13561] issue: Custom Modell slowness #87693

New Issue

2026-05-15T12:10:45-05:00

GiteaMirror commented

2026-05-15 12:10:45 -05:00

Originally created by @Xeroxxx on GitHub (May 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/13561

Check Existing Issues

I have searched the existing issues and discussions.
I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.6.6

Ollama Version (if applicable)

v0.6.8 (CUDA; Quadro P1000)

Operating System

Debain Trixie - Kubernetes

Browser (if applicable)

Edge 137.0.3282.1

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 listed steps to reproduce the bug in detail.

Expected Behavior

Hello,
I'm running on kubernetes and created a custom model. Its based on LLama3.2. There is no addtional plugins or knowledge. Just a system prompt, no extras.

When using LLama3.2 the response starts is instant. When I use the custom model with the system prompt the response takes up to one minute. I can't see any CPU or GPU utilization till it starts. What am I doing wrong here?

I can see the prompt is pushed to ollama round about after a minute.

Thanks!

Actual Behavior

Custom Modells response takes up to one minute to start.

Steps to Reproduce

Create a custom modell based on llama3.2
Add a system prompt
Ask AI a Quesiton
Wait a minute
Ask llama3.2 modell
get instant response

Logs & Screenshots

Stats from custom modell response:

Additional Information

I'm out of ideas whats wrong here.

Originally created by @Xeroxxx on GitHub (May 6, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/13561 ### Check Existing Issues - [x] I have searched the existing issues and discussions. - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version v0.6.6 ### Ollama Version (if applicable) v0.6.8 (CUDA; Quadro P1000) ### Operating System Debain Trixie - Kubernetes ### Browser (if applicable) Edge 137.0.3282.1 ### 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 listed steps to reproduce the bug in detail. ### Expected Behavior Hello, I'm running on kubernetes and created a custom model. Its based on LLama3.2. There is no addtional plugins or knowledge. Just a system prompt, no extras. When using LLama3.2 the response starts is instant. When I use the custom model with the system prompt the response takes up to one minute. I can't see any CPU or GPU utilization till it starts. What am I doing wrong here? I can see the prompt is pushed to ollama round about after a minute. Thanks! ### Actual Behavior Custom Modells response takes up to one minute to start. ### Steps to Reproduce 1. Create a custom modell based on llama3.2 2. Add a system prompt 3. Ask AI a Quesiton 4. Wait a minute 5. Ask llama3.2 modell 6. get instant response ### Logs & Screenshots ![Image](https://github.com/user-attachments/assets/b2872650-883a-4d88-a324-29a36bc19f6e) Stats from custom modell response: ![Image](https://github.com/user-attachments/assets/b455d667-0e90-4978-8957-940e6a559a2f) ### Additional Information I'm out of ideas whats wrong here.

GiteaMirror added the bug label 2026-05-15 12:10:45 -05:00

GiteaMirror closed this issue

2026-05-15 12:10:47 -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#87693