openwebui censorship? #2620

New Issue

2025-11-11T15:10:50-06:00

GiteaMirror commented

2025-11-11 15:10:50 -06:00

Originally created by @joshuacox on GitHub (Nov 12, 2024).

Installation Method

docker compose

Environment

Open WebUI Version: [e.g., v0.3.11]
Ollama (if applicable): [e.g., v0.2.0, v0.1.32-rc1]

both installed from the docker compose:

commit 96c865404d36637eafadb6d2dd2365c85d452648
Author: Timothy Jaeryang Baek <tim@openwebui.com>
Date:   Sat Nov 9 18:01:23 2024 -0800

    refac: disable openapi.json

Operating System: [e.g., Windows 10, macOS Big Sur, Ubuntu 20.04]

cat /etc/os-release
PRETTY_NAME="Ubuntu 22.04.5 LTS"
NAME="Ubuntu"
VERSION_ID="22.04"
VERSION="22.04.5 LTS (Jammy Jellyfish)"
VERSION_CODENAME=jammy
ID=ubuntu
ID_LIKE=debian
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
UBUNTU_CODENAME=jammy

Browser (if applicable): [e.g., Chrome 100.0, Firefox 98.0]

Version 130.0.6723.116 (Official Build) Arch Linux (64-bit)

Confirmation:

I have read and followed all the instructions provided in the README.md.
I am on the latest version of both Open WebUI and Ollama.

I'll not include the logs, as that would expose my users and what they asked, you can test this yourself with whatever offensive prompt that you can dream up.

Expected Behavior:

I expect openwebui to pass my users prompt to the model without censorship.

Actual Behavior:

openwebui censors my users prompt, whereas I can run the model directly with ollama run model and it answers my prompt truthfully.

Description

Censorship in openwebui

Bug Summary:

OpenWebui appears to be censoring input.

Reproduction Details

Steps to Reproduce:

find a prompt offensive enough to get rejected using an uncensored model like: wizard-vicuna-uncensored

then run the same prompt using ollama run wizard-vicuna-uncensored

Logs and Screenshots

Again, the prompts and screenshots themselves would obviously be inappropriate for github. But I maintain that hosting a private instance of openwebui, and running uncensored LLMs, that openwebui should not be interjecting itself, the LLM should speak for itself.

Originally created by @joshuacox on GitHub (Nov 12, 2024). ## Installation Method docker compose ## Environment - **Open WebUI Version:** [e.g., v0.3.11] - **Ollama (if applicable):** [e.g., v0.2.0, v0.1.32-rc1] both installed from the docker compose: ``` commit 96c865404d36637eafadb6d2dd2365c85d452648 Author: Timothy Jaeryang Baek <tim@openwebui.com> Date: Sat Nov 9 18:01:23 2024 -0800 refac: disable openapi.json ``` - **Operating System:** [e.g., Windows 10, macOS Big Sur, Ubuntu 20.04] cat /etc/os-release PRETTY_NAME="Ubuntu 22.04.5 LTS" NAME="Ubuntu" VERSION_ID="22.04" VERSION="22.04.5 LTS (Jammy Jellyfish)" VERSION_CODENAME=jammy ID=ubuntu ID_LIKE=debian HOME_URL="https://www.ubuntu.com/" SUPPORT_URL="https://help.ubuntu.com/" BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/" PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy" UBUNTU_CODENAME=jammy - **Browser (if applicable):** [e.g., Chrome 100.0, Firefox 98.0] Version 130.0.6723.116 (Official Build) Arch Linux (64-bit) **Confirmation:** - [x] I have read and followed all the instructions provided in the README.md. - [x] I am on the latest version of both Open WebUI and Ollama. I'll not include the logs, as that would expose my users and what they asked, you can test this yourself with whatever offensive prompt that you can dream up. ## Expected Behavior: I expect openwebui to pass my users prompt to the model without censorship. ## Actual Behavior: openwebui censors my users prompt, whereas I can run the model directly with `ollama run model` and it answers my prompt truthfully. ## Description Censorship in openwebui **Bug Summary:** OpenWebui appears to be censoring input. ## Reproduction Details **Steps to Reproduce:** find a prompt offensive enough to get rejected using an uncensored model like: wizard-vicuna-uncensored then run the same prompt using `ollama run wizard-vicuna-uncensored` ## Logs and Screenshots Again, the prompts and screenshots themselves would obviously be inappropriate for github. But I maintain that hosting a private instance of openwebui, and running uncensored LLMs, that openwebui should not be interjecting itself, the LLM should speak for itself.

GiteaMirror closed this issue

2025-11-11 15:10:50 -06:00

GiteaMirror commented

2025-11-11 15:10:51 -06:00