[GH-ISSUE #15143] issue: When using Azure AI Speech for TTS. It does not honour the user set voice defaulting to the admin set value #72088

New Issue

GiteaMirror · 2026-05-13T04:05:53-05:00

GiteaMirror commented

2026-05-13 04:05:53 -05:00

Originally created by @Dudedrakes on GitHub (Jun 19, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15143

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

v0.6.15

Ollama Version (if applicable)

No response

Operating System

All

Browser (if applicable)

Chrome and Edge

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 using Azure AI Speech, the user should be able to override the admin-set voice and pick their own from the list of available voices. Once set in the user's personal settings and saved, all subsequent read-aloud requests and Voice Mode sessions should then use the user's selected voice, just like it does when using OpenAI as the TTS provider.

Actual Behavior

The actual behavior is that the read-aloud and voice mode sessions are ignoring the user set voice and defaulting to the admin set value even if the user set value is different to the admin set value.

Steps to Reproduce

These instructions assume you already have a functional Open-WebUI installation. I have tried this on multiple instances running in Docker containers across various environments with multiple models via an OpenAI-compatible endpoint. The problem exists whether you are using a private endpoint with a custom domain or not.

Deploy an Azure AI Speech Instance in Azure
Go to the audio settings in the admin panel in Open Web-UI
Configure TTS to use Azure AI Speech using the details from the instance you've just deployed. Correctly setting the region or endpoint, voice and output format.
Go to an existing chat or start a new one.
Use either Voice mode or read aloud to get Open Web-UI to read out a response
Go to the audio panel in user settings
Change the set voice to a different value and click save
Go to an existing chat or start a new one.
Use either Voice mode or read aloud to get Open Web-UI to read out a response

You'll notice the voices are the same despite the user-set value being different.

Logs & Screenshots

The app logs do not show anything erroneous

The browser console has no entries.

Here is my admin panel config

Here is a user setting

Here is the speech request in the browser showing that the user-selected voice is submitted to the speech API endpoint

Additional Information

No response

Originally created by @Dudedrakes on GitHub (Jun 19, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/15143 ### 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 v0.6.15 ### Ollama Version (if applicable) _No response_ ### Operating System All ### Browser (if applicable) Chrome and Edge ### 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 using Azure AI Speech, the user should be able to override the admin-set voice and pick their own from the list of available voices. Once set in the user's personal settings and saved, all subsequent read-aloud requests and Voice Mode sessions should then use the user's selected voice, just like it does when using OpenAI as the TTS provider. ### Actual Behavior The actual behavior is that the read-aloud and voice mode sessions are ignoring the user set voice and defaulting to the admin set value even if the user set value is different to the admin set value. ### Steps to Reproduce These instructions assume you already have a functional Open-WebUI installation. I have tried this on multiple instances running in Docker containers across various environments with multiple models via an OpenAI-compatible endpoint. The problem exists whether you are using a private endpoint with a custom domain or not. 1. Deploy an Azure AI Speech Instance in Azure 2. Go to the audio settings in the admin panel in Open Web-UI 3. Configure TTS to use Azure AI Speech using the details from the instance you've just deployed. Correctly setting the region or endpoint, voice and output format. 4. Go to an existing chat or start a new one. 5. Use either Voice mode or read aloud to get Open Web-UI to read out a response 6. Go to the audio panel in user settings 7. Change the set voice to a different value and click save 8. Go to an existing chat or start a new one. 9. Use either Voice mode or read aloud to get Open Web-UI to read out a response You'll notice the voices are the same despite the user-set value being different. ### Logs & Screenshots The app logs do not show anything erroneous ![Image](https://github.com/user-attachments/assets/2066f03b-4009-4d52-9cff-77bf8a6b6618) The browser console has no entries. Here is my admin panel config ![Image](https://github.com/user-attachments/assets/fd03e038-464e-4d71-b83d-eb57a0844a50) Here is a user setting ![Image](https://github.com/user-attachments/assets/e254749c-37ed-4b07-9514-d5111888c4ce) Here is the speech request in the browser showing that the user-selected voice is submitted to the speech API endpoint ![Image](https://github.com/user-attachments/assets/8d921707-8dce-41a9-a969-a8236629bdc5) ### Additional Information _No response_

GiteaMirror added the bug label 2026-05-13 04:05:53 -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#72088